U.S. patent application number 12/468879 was filed with the patent office on 2009-12-03 for vehicle pre-impact sensing system having object feature detection.
This patent application is currently assigned to Delphi Technologies, Inc.. Invention is credited to Kevin J. Hawes, Ronald M. Taylor.
Application Number | 20090299631 12/468879 |
Document ID | / |
Family ID | 41021330 |
Filed Date | 2009-12-03 |
United States Patent
Application |
20090299631 |
Kind Code |
A1 |
Hawes; Kevin J. ; et
al. |
December 3, 2009 |
Vehicle Pre-Impact Sensing System Having Object Feature
Detection
Abstract
A vehicle pre-impact sensing system is provided that includes an
array of energy signal transmitters mounted on a vehicle for
transmitting signals within multiple transmit zones spaced from the
vehicle and an array of receiver elements mounted on the vehicle
for receiving signals reflected from an object located in one or
more multiple receive zones indicative of the object being in
certain one or more zones. A processor processes the received
reflected signals and determines range, location, speed and
direction of the object, determines whether the object is expected
to impact the vehicle as a function of the determined range,
location, speed and direction of the object, and generates an
output signal indicative of a pre-impact event. The system may
detect one or more features of a target object, such as a front end
of a vehicle. Additionally, the system may modulate the transmit
beams. Further, the system may perform a terrain normalization to
remove stationary items.
Inventors: |
Hawes; Kevin J.; (Greentown,
IN) ; Taylor; Ronald M.; (Greentown, IN) |
Correspondence
Address: |
DELPHI TECHNOLOGIES, INC.
M/C 480-410-202, PO BOX 5052
TROY
MI
48007
US
|
Assignee: |
Delphi Technologies, Inc.
Troy
MI
|
Family ID: |
41021330 |
Appl. No.: |
12/468879 |
Filed: |
May 20, 2009 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61130236 |
May 29, 2008 |
|
|
|
Current U.S.
Class: |
701/300 ;
250/338.1; 250/340; 359/350 |
Current CPC
Class: |
G01S 7/4802 20130101;
G01S 17/87 20130101; G01S 17/931 20200101; B60R 2021/01327
20130101; G01S 2013/93274 20200101; B60R 2021/0006 20130101; B60R
21/0134 20130101; B60R 2021/01027 20130101 |
Class at
Publication: |
701/300 ;
359/350; 250/338.1; 250/340 |
International
Class: |
G06F 19/00 20060101
G06F019/00; G06K 7/10 20060101 G06K007/10; G01J 5/00 20060101
G01J005/00; G01J 5/02 20060101 G01J005/02 |
Claims
1. A vehicle pre-impact sensing system comprising: an array of
energy signal transmitters mounted on a vehicle for transmitting
signals within multiple transmit zones spaced from the vehicle; an
array of receiver elements mounted on the vehicle for receiving the
signals reflected from an object located in one or more multiple
receive zones indicative of the object being in certain one or more
receive zones; and a processor for processing the received
reflected signals and determining range, location, speed and
direction of the object, wherein the processor processes the
received reflected signals to detect a feature of the object and
determines the range at least partially based on the detected
feature, and wherein the processor further determines whether the
object is expected to impact the vehicle as a function of the
determined range, location, speed and direction of the object, and
generates an output signal indicative of a sensed pre-impact
event.
2. The sensing system as defined in claim 1, wherein the array of
energy signal transmitters comprises an array of light
transmitters.
3. The sensing system as defined in claim 2, wherein the array of
light transmitters comprises an array of infrared transmitters
configured to emit infrared radiation in designated multiple
transmit zones.
4. The sensing system as defined in claim 3, wherein the array of
receiver elements comprises an array of photodetectors for
receiving reflected infrared radiation, wherein the photodetectors
receive light signals including reflected signals from designated
receive zones.
5. The sensing system as defined in claim 1, wherein the system
senses an object on a lateral side of the vehicle.
6. The sensing system as defined in claim 5, wherein the array of
receiver elements are mounted on an exterior mirror housing.
7. The sensing system as defined in claim 1, wherein the processor
further determines size of the object, and wherein the processor
generates an output signal indicative of a sensed pre-impact event
further based on the determined size of the object.
8. The sensing system as defined in claim 1 further comprising a
horizontally illuminated signal transmitter projecting a
horizontally illuminated projected beam elevated at a height to
identify a feature on the object, wherein the array of receiver
elements receive signals reflected objects in the horizontal beam
and multiplex the horizontal beam signal with the array of receiver
signals.
9. The sensing system as defined in claim 8, wherein the processor
determines a reflection coefficient of the surface of the object
and determines range based at least in part on the reflection
coefficient.
10. The sensing system as defined in claim 1, wherein the array of
energy signal transmitters comprises an array of shaped IR
transmitters each having a beam shape representative of a known
feature of the object.
11. The sensing system as defined in claim 10, wherein the known
feature comprises a feature found on a common vehicle front
side.
12. The sensing system as defined in claim 11, wherein the beam
shape has an approximate U-shape.
13. The sensing system as defined in claim 1, wherein the processor
further detects a differential spectral return of highly reflective
surfaces of an object and further determines range to the object
based on the detected differential spectral return.
14. The sensing system as defined in claim 1, wherein the processor
detects reflective vehicle surfaces comprising one or more of a
signal marker, a headlamp, a fog lamp, a license plate and a chrome
feature of a vehicle.
15. The sensing system as defined in claim 14, wherein background
light illumination is used to measure highly reflective vehicle
elements.
16. The sensing system as defined in claim 15, wherein the
processor further detects a headlamp on status of an inbound
vehicle and further determines range based on the detected
headlamp.
17. The sensing system as defined in claim 16, wherein the
processor determines a geometry of the detected object and further
employs the geometry in the discrimination risk assessment.
18. A method of sensing pre-impact with a vehicle, said method
comprising the steps of: transmitting energy signals with an array
of energy transmitters within multiple transmit zones spaced from
the vehicle; receiving energy signals including signals reflected
from an object located in the one or more multiple receive zones
with an array of receiver elements; processing the received
reflected signals and determining range, location, speed and
direction of the object; processing the received reflected signals
to detect a feature of the object; determining the range at least
partially based on the detected feature; determining whether the
object is expected to impact the vehicle as a function of the
determined range, location, speed and direction of the object; and
generating an output signal indicative of a sensed pre-impact
event.
19. The method as defined in claim 18, wherein the step of
transmitting signals comprises transmitting an infrared radiation
signal via an array of infrared transmitters into the multiple
transmit zones, one zone at a time, and receiving the signals
comprising received reflected infrared radiation via an array of
infrared receivers, one receiver at a time.
20. The method as defined in claim 18, wherein the array of
transmitters and receivers are oriented generally on a lateral side
direction of the vehicle to detect an object toward the lateral
side of the vehicle.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application claims priority under 35 U.S.C.
.sctn.119(e) to U.S. Provisional Patent Application No. 61/130,236,
filed on May 29, 2008, the entire disclosure of which is hereby
incorporated herein by reference.
TECHNICAL FIELD
[0002] The present application generally relates to vehicle crash
sensing and, more particularly, relates to a system and method of
sensing an imminent collision of an object with a vehicle prior to
impact.
BACKGROUND OF THE INVENTION
[0003] Automotive vehicles are commonly equipped with passenger
restraint and crash mitigation devices such as seat belts, front
air bags, side air bags and side curtains. These and other devices
may be deployed in the event of a collision with the host vehicle
to mitigate adverse effects to the vehicle and the occupants in the
vehicle. With respect to activated devices, such as air bags and
side curtain bags, these devices generally must be deployed quickly
and in a timely fashion. Typically, these types of devices are
deployed when sensors (e.g., accelerometers) mounted on the vehicle
sense a severe impact with the vehicle.
[0004] In some vehicle driving situations, it is desirable to
determine the onset of a collision, prior to impact of an object
with the host vehicle. For example, vision systems employing
cameras may be used to monitor the surrounding environment around
the vehicle and the video images may be processed to determine if
an object appears to be on a collision path with the vehicle.
However, visions systems are generally very expensive and suffer a
number of drawbacks.
[0005] An alternative approach is disclosed in U.S. Patent
Application Publication No. 2009/0099736, assigned to the assignee
of the present application. The approach set forth in the
aforementioned patent application discloses a vehicle pre-impact
sensing system that transmits a plurality of infrared (IR) beams
and receives a plurality of beams within a plurality of curtains
incrementally spaced from the host vehicle for sensing objects that
may impact the side of the host vehicle. The aforementioned
published patent application is hereby incorporated herein by
reference.
[0006] It would be desirable to provide for an enhanced
cost-effective system that senses a collision prior to impact with
the host vehicle, particularly for use to detect side impact
events.
SUMMARY OF THE INVENTION
[0007] According to one aspect of the present invention, a vehicle
pre-impact sensing system is provided that includes an array of
energy signal transmitters mounted on a vehicle for transmitting
signals within multiple transmit zones spaced from the vehicle. The
system further includes an array of receiver elements mounted on
the vehicle for receiving signals reflected from an object located
in one or more multiple receive zones indicative of the object
being in certain one or more receive zones. The system also
includes a processor for processing the received reflected signals
and determining range, location, speed and direction of the object.
The processor processes the received reflected signals to detect a
feature of the object and determines the range at least partially
based on the detected feature. The processor further determines
whether the object is expected to impact the vehicle as a function
of the determined range, location, speed and direction of the
object, and generates an output signal indicative of a sensed
pre-impact event.
[0008] According to one aspect of the embodiment of the present
invention, a method of sensing per-impact with a vehicle is
provided. The method includes the steps of transmitting energy
signals with an array of energy transmitters within multiple
transmit zones spaced from the vehicle. The method further includes
receiving energy signals including signals reflected from an object
located in the one or more multiple receive zones with an array of
receiver elements, processing the received reflected signals and
determining range, location, speed and direction of the object and
processing the received reflected signals to detect a feature of
the object. The method also determines the range at least partially
based on the detected feature and determines whether the object is
expected to impact the vehicle as a function of the determined
range, location, speed and direction of the object. The method
further generates an output signal indicative of a sensed
pre-impact event.
[0009] These and other features, advantages and objects of the
present invention will be further understood and appreciated by
those skilled in the art by reference to the following
specification, claims and appended drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] The present invention will now be described, by way of
example, with reference to the accompanying drawings, in which:
[0011] FIG. 1 is a side perspective view of a vehicle employing a
pre-impact crash sensing system illustrating an array of infrared
(IR) transmit zones, according to one embodiment;
[0012] FIG. 2 is a side perspective view of the vehicle employing
the pre-impact crash sensing system showing an array of receiver
photodetection zones, according to one embodiment;
[0013] FIG. 3 is a top view of the vehicle further illustrating the
IR transmit zones employed in the crash sensing system of FIG.
1;
[0014] FIG. 4 is a top view of the vehicle further illustrating the
IR photoreceiver zones employed in the crash sensing system shown
in FIG. 2;
[0015] FIG. 5 is an enlarged view of an integrated IR
transmitter/receiver employed in the crash sensing system,
according to one embodiment;
[0016] FIG. 6 is a block diagram illustrating the pre-impact crash
sensing system, according to one embodiment;
[0017] FIG. 7 is a flow diagram illustrating a routine for sensing
a pre-impact collision of an object with the vehicle, according to
one embodiment;
[0018] FIG. 8 is a rear view of the vehicle showing the IR transmit
zones including an additional elevated horizontal IR transmit zone,
according to one embodiment;
[0019] FIGS. 9A and 9B is a flow diagram illustrating a routine for
sensing a vehicle feature to update a range estimate and a side
impact crash, according to one embodiment;
[0020] FIG. 10 is a calibration chart illustrating reflected ripple
signals as a function of distance for various identified objects,
according to one example;
[0021] FIG. 11 illustrates a U-shaped IR transmit beam superimposed
onto the front of an oncoming vehicle for use in detecting a
vehicle fascia feature(s), according to one embodiment;
[0022] FIG. 12 illustrates an array of the U-shaped IR transmit
beams shown in FIG. 11 superimposed onto an oncoming vehicle for
detecting the vehicle fascia feature(s);
[0023] FIG. 13 illustrates receiver output signals and ripple
signal in frequency (Hz) as a function of time as the host vehicle
drives by another vehicle, according to one example;
[0024] FIG. 14 illustrates the host vehicle traveling relative to a
stationary barrier, lateral displaced vehicles and a lateral
approaching vehicle, to illustrate terrain normalization, according
to one embodiment;
[0025] FIGS. 15A-15E illustrate the passing of a stationary object
and terrain normalization to detect the object as stationary;
[0026] FIGS. 15A-A-15E-E are timing diagrams that illustrate
normalization of a detected object as it passes through detection
zones A1-A3 shown in FIGS. 15A-15E, respectively;
[0027] FIGS. 16A-16D illustrate the passing of an angled stationary
object and the terrain normalization for detecting the stationary
target;
[0028] FIGS. 16A-A-16D-D are timing diagrams illustrating terrain
normalization as the angled object passes through the detection
zones shown in FIGS. 16A-16D, respectively;
[0029] FIGS. 17A-17D illustrate terrain normalization on an object
moving laterally with respect to the host vehicle, according to one
example; and
[0030] FIGS. 18A-18C illustrate a routine for providing terrain
normalization, according to one embodiment.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0031] Referring to FIGS. 1-5, a vehicle pre-impact crash sensing
system 20 is generally illustrated employed on a host vehicle 10,
according to one embodiment. The crash sensing system 20 is shown
and described herein configured to detect a pre-impact collision of
an object (e.g., another vehicle) with the vehicle 10, particularly
on one or both lateral sides of the vehicle 10. However, it should
be appreciated that the crash sensing system 20 may be employed to
detect a pre-impact event on any side of the vehicle 10, including
one or both lateral sides, the front side and the rear side.
[0032] The host vehicle 10 is generally shown as an automotive
wheeled vehicle having opposite lateral sides and exterior side
view mirror housings 12 on opposite lateral sides. In the
embodiment shown and described herein, the crash sensing system 20
generally includes an integrated infrared (IR) transmitter/receiver
25 shown mounted generally in one of the mirror housings 12 of the
vehicle 10, at a position sufficient to detect objects located
adjacent to the corresponding lateral side of the vehicle 10. While
lateral crash sensing is shown and described herein for sensing a
collision on one side of the host vehicle 10, it should be
appreciated that the crash sensing may also be employed on the
opposite lateral side of the vehicle. Further, while the
transmitter/receiver 25 is shown mounted in the mirror housing 12,
it should be appreciated that the integrated transmitter/receiver
array 25 may be located at other locations on the vehicle 10 and
positioned to detect one or more objects in the desired vicinity of
the vehicle 10.
[0033] The IR transmitter/receiver 25 includes a plurality of IR
transmitters (22A-22I) and a plurality of IR receivers 24A-24I, as
shown in FIG. 5. As seen in FIGS. 1 and 3, the IR transmitters
22A-22I transmit infrared (IR) energy signals within designated
transmit beam patterns 32A-32I in a three-by-three (3.times.3)
array spaced from the lateral side of the host vehicle 10,
according to one embodiment. The infrared transmitter array 25 has
a plurality (e.g., nine) of infrared transmitters 22A-22I for
transmitting infrared radiation signals within designated
corresponding transmit zones 32A-32I. The transmitter array is
activated to sequentially transmit infrared radiation signals in
one zone at a time, such as zone 32A, and then switches
sequentially to the next zone, such as zone 32B, and then to zone
32C, and continues through the entire array to zone 32I, and
cyclically repeats the process at a high rate of speed, e.g., less
than three milliseconds per zone. Alternately, multiple transmit
zones could be illuminated simultaneously. In the embodiment shown,
the IR transmit zones 32A-32I are oriented in a three-by-three
(3.times.3) array having three rows and three columns, each zone
having a generally conical shape extending from the
transmitter/receiver 25 shown located in the mirror housing 12 and
oriented toward the roadway to the lateral side of the host vehicle
10 such that the row of zones 32A-32C is spaced further away from
the host vehicle 10 as compared to the row of zones 32D-32F and row
of zones 32G-32I. Each IR transmit zone 32A-32I has a generally
cone shape beam with a circular cross section which appears more as
an elliptical shape as it impinges at an angle on the ground on the
adjacent roadside. Each IR transmit zone 32A-32I has a size
sufficient to cover the intended detection zone at the lateral side
of the host vehicle 10 and may be spaced from the adjacent zones by
a predetermined angle, according to one embodiment. According to
another embodiment, the IR transmit zones 32A-32I may overlap each
other, thereby offering intermediary zones that can be further
processed.
[0034] The crash sensing system 20 also includes a receiver array
having a plurality of photosensitive receiver elements 24A-24I as
shown in FIG. 5. In one embodiment, the receiver elements 24A-24I
receive and detects light intensity including reflected infrared
radiation signals within corresponding IR receive zones 34A-34I as
shown in FIGS. 2 and 4. The IR receivers 24A-24I essentially
receive light including infrared radiation signals reflected from
one or more objects within the corresponding IR receive zones
34A-34I and converts the detected light intensity to a frequency
output. In one embodiment, the receive zones 34A-34I are arranged
in a three-by-three (3.times.3) array having three columns and
three rows that are located such as to substantially align with the
corresponding three-by-three (3.times.3) array of IR transmit zones
32A-32I.
[0035] In addition, the vehicle 10 is shown in FIG. 1 having an
additional or tenth IR transmitter 26 shown illuminating a
horizontal beam at an elevation where an oncoming vehicle bumper or
grille would be expected to be located. The additional IR
transmitter 26 transmits a substantially horizontal IR calibration
beam outward from the lateral side direction of the host vehicle 10
as shown in FIG. 8. In the embodiment shown, IR transmitter 26 is
located in a passenger door of the host vehicle 10 at a height
similar to or the same as the elevation of the beam 28 relative to
the ground. It should be appreciated that the transmitter 26 may be
located elsewhere on host vehicle 10 such as in the front quarter
panel or the front or rear bumpers of host vehicle 10. The
additional IR transmit beam 28 provides a horizontal calibration IR
beam which is illuminated by itself in the timing sequence of
illuminating the nine IR transmitters 22A-22I. By illuminating the
additional IR transmitter 26, the IR receivers 24A-24I may detect
light intensity including infrared radiation reflected from objects
in the corresponding zones 34A-34I, particularly for object
features located at the elevation of beam 28 such as a vehicle
bumper, a vehicle grille or fascia or other features expected to be
detected at that elevation. By providing the extra horizontal IR
transmit beam 28, triangulation of the calibration beam with the
other nine scanned IR beams 32A-32I allows ranging and a measure of
the reflection coefficient of the surface of a target object. As
such, enhanced range information may be acquired, according to one
embodiment.
[0036] Further, the host vehicle 10 is shown in FIG. 2 having an
optional additional IR photoreceiver 27 shown located at the same
or in close proximity to IR transmitter 26 for receiving reflected
signals within a horizontal beam at an elevation where an oncoming
vehicle bumper or grille would be expected to be located. The IR
receiver 27 receive light signals including reflected IR signals
within receive beam 29 and generates a frequency output as a
function of the detected light amplitude. It should be appreciated
that with the additional IR transmitter 26 turned on, either the
additional IR photoreceiver 27 or the individual IR photoreceivers
24A-24I may be employed to detect the presence of an object within
the horizontal beams at the elevation where an oncoming vehicle
bumper or grille is expected which enables enhanced range
estimation based on triangulation of the received signals.
[0037] In operation, the array of IR transmitters 22A-22I transmits
infrared radiation signals within the corresponding IR transmit
zones 32A-32I, one zone at a time, according to one embodiment,
resulting in the transmission of sequential IR signals to the
transmit zones, while the receiver array 24A-24I receives light
energy including reflected infrared radiation signals from objects
located within the corresponding IR receive zones 34A-34I. The
detected light signals are output as frequency signals which are
then processed by a processor. By knowing which one of the IR
transmit zones 32A-32I is illuminated with infrared radiation at a
given point in time, the location and range of the detected object
can be determined. As an object moves, the progression of the
object through multiple zones can be monitored to determine speed
and direction of the object, such that a processor may determine
whether a pre-impact event of the object with the host vehicle 10
is detected. In addition to the sequential illumination of IR
transmitters 22A-22I, the system 20 may also activate the
additional IR transmitter 26 as part of the sequence to detect
objects within the illuminated horizontal IR calibration beam 28.
The sequence of illumination may include successive activation of
the nine IR transmitters 22A-22I, the activation of the tenth IR
transmitter 26, then all IR transmitters turned off, and then
repeat the cycle.
[0038] With particular reference to FIG. 5, the integrated IR
transmitter/receiver 25 is illustrated, according to one
embodiment. The IR transmitter/receiver 25 is shown generally
having a housing 60 containing an array of nine IR LEDs 22A-22I
mounted onto the top side of a circuit board 62. The IR LEDs
22A-22I may be disposed behind respective beam-forming optics,
which may include reflecting (e.g., parabolic reflector) and/or
refracting optical elements or an aperture for defining the
conical-shaped beam pattern. It should be appreciated that the IR
transmitter array 22A-22I may employ any of a number of signal
transmitting elements for illuminating multiple transmit zones, and
may be configured in any of a number of shaped and sized beam
patterns. According to one example, the IR LEDs 22A-22I may employs
a central wavelength of about 850 nanometers. One example of a
commercially available IR LED is available from OSRAM Opto
Semiconductors Inc., sold under the brand name Golden Dragon.
[0039] The IR transmitter/receiver 25 is shown employing nine
photodetectors 24A-24I which serve as photosensitive receivers and
are shown mounted on the bottom side of the circuit board 62.
Photodetectors 24A-24I may generally be placed behind corresponding
receiving lenses and/or receiving reflectors (e.g., parabolic
reflectors). The receiving lenses may include reflecting and/or
refracting optical elements that focus the reflected infrared
radiation received from the corresponding IR receive zones 34A-34I
onto the photodetectors 24A-24I, respectively. The receiver array
may employ any number of a plurality of receiver elements for
receiving reflected IR signals from objects within the
corresponding number of receive zones 24A-24I and may each be
configured in a cone shape or other shapes and sizes. One example
of a photodetector is a light-to-frequency converter commercially
available from Texas Advanced Optoelectronic Solutions (TAOS). The
light-to-frequency converter provides a frequency output (Hz) as a
function of the amplitude of the detected light radiation.
[0040] Referring to FIG. 6, the crash sensing system 20 is further
illustrated employing a microprocessor 50 having various inputs and
outputs. The microprocessor 50 is shown providing outputs to the
nine IR transmitters 22A-22I and the tenth IR transmitter 26. The
microprocessor 50 outputs LED strobe signals to the IR LEDs 22A-22I
and 26 to activate the IR LEDs in a cyclical pattern. Signals
indicating light and reflected infrared radiation received by each
of the receiver elements 24A-24I are input to the microprocessor
50. In the embodiment disclosed, the receiver elements 24A-24I
provide frequency signals input to the microprocessor 50 in the
form of a beam IR signature. The beam IR signature includes
frequency (in hertz) representing the amplitude of the photo or
light energy detected by its irradiance on the receiver from within
a given detection zone.
[0041] In addition, the microprocessor 50 receives an input from a
passive thermal far IR receiver 46. The passive thermal far IR
receiver 46 detects emitted radiation within a relatively large
area and serves as a safing input that may be logically ANDed with
a processor generated output signal to provide risk mitigation for
high target certainty. Alternately, the crash sensing system 20 may
employ radar or an ultrasonic transducer as the safing input.
Further safing inputs may include a steering wheel angular rate,
yaw, external lateral slip, lateral acceleration and lateral
velocity signals, amongst other possible safing inputs.
[0042] The crash sensing system 20 further includes memory 52,
including volatile and/or non-volatile memory, which may be in the
form of random access memory (RAM), electrically erasable
programmable read-only memory (EEPROM) or other memory storage
medium. Stored within the memory 52 is a sensing routine 100 for
processing the sensed data and determining a pre-impact event as
described herein. Also stored in memory 52 and executed by
microprocessor 50 is a routine 200 detecting vehicle features and
determining a pre-impact event and a routine 300 performing terrain
normalization and determining a pre-impact event.
[0043] Additionally, the microprocessor 50 provides a resettable
countermeasure deploy output signal 54 and a non-resettable
countermeasure deploy output signal 56. The countermeasure deploy
output signals 54 and 56 may be employed to mitigate the effects of
an anticipated collision. Examples of countermeasure deploy
activity may include deploying a pretensioner for one or more seat
belts, deploying one or more air bags and/or side air bag curtains,
controlling an active suspension or other vehicle dynamics
adjustment, or further may activate other countermeasures on board
the host vehicle 10. These are other deployments may be initiated
early on, even prior to an actual impact. Further, the
microprocessor 50 receives vehicle speed 58 which may be a measured
vehicle speed on a vehicle speed estimate. Vehicle speed 58 may be
employed to determine whether or not a lateral impact with an
object is expected and is further employed for terrain
normalization to determine whether or not an object is stationary
despite its shape and orientation.
[0044] The sensing routine 100 is illustrated in FIG. 7 for sensing
an anticipated near impact event of an object with the host
vehicle. Routine 100 begins at step 102 to sequentially transmit IR
beams, and the proceeds to step 104 to monitor received
photosensitive beams within the array of coverage zones. This
occurs by sequentially applying IR radiation within each of the
transmit beams and receiving light energy including reflected IR
signals from the receive zones. Next, routine 100 performs noise
rejection on the beam data that is received. In decision step 108,
routine 100 determines if the temporal gating has been met and, if
not, returns to step 102. The temporal gating bracketing
requirements may take into consideration the path, trajectory and
rate of the object, the number of pixels of area, the inferred
mass/volume/area, the illumination consistency, and angular beam
spacing consistency.
[0045] According to one embodiment, temporal gating requirements
are determined based on comparison of an object's perceived motion
(detection from one contiguous coverage zone to another) across the
coverage zones to the expected relative speed of potential
collision objects of interest (e.g., an automotive vehicle moving
at a closing speed of 10 to 65 kilometers per hour (kph) or 6 to 40
miles per hour (mph) to a host vehicle's lateral side). The "range
rate" of distance traveled per unit time of a potential collision
object can be determined by the detection assessment of contiguous
coverage zones for range rates consistent with an expected subject
vehicle's closing speed (i.e., if an object is detected passing
through the coverage zones at a rate of 1 observation zone per 70
milliseconds and each coverage zone is 0.3 meters in diameter
perpendicular to the host vehicle's lateral side, then the closing
speed or range rate is approximately 4 meters per second, and is
equivalent to approximately 15 kph or 10 mph). Objects moving at
range rates slower or faster than the expected range rate boundary
through the coverage zones would not pass the temporal gating
requirement.
[0046] Additional assessment can be made based on the quality of
the received signal of a potential object as it passes through the
coverage zones. If the amplitude of the detected signal varies
substantially from one contiguous coverage zone to another (even if
all signals are above a threshold value), it could indicate an
off-axis collision trajectory or perhaps an object with a mass not
consistent with a vehicle. The signal fidelity and consistency
through the contiguous coverage zones can be used to verify a
potential vehicle collision.
[0047] If the temporal gating has been met, routine 100 then
proceeds to decision step 110 to determine if the far IR safing has
been enabled and, if not, returns to step 102. If the safing has
been enabled, routine 100 proceeds to deploy an output signal
indicative of a sensed pre-impact event in step 112. The output
signal may be employed to activate deployment of one or more
countermeasures.
[0048] The crash sensing system 20 creates a three-dimensional
space extending from the lateral side of the host vehicle 10 by way
of an array of high speed sequentially illuminated and scanned
infrared light signals provided in dedicated coverage zones
directed to the lateral side of the host vehicle 10. Objects which
appear within the coverage zones are scanned, and their location,
range, speed, and direction are determined. In addition, the size
of the object may be calculated. Further, the shape of the object
and one or more features such as reflectivity present on the object
may further be further determined. It should be appreciated that
feature identification, such as may be achieved by monitoring
reflectivity, such as that due to color, and other variations, may
be detected and an enhanced range may be determined. The processor
50 processes the information including location, range, speed and
direction of the object in addition to the host vehicle speed, and
determines whether or not a detected object is expected to impact
the side of the host vehicle 10. The processor 50 processes the
location of the detected object, range to the detected object,
speed of the detected object, and direction of the detected object
in relation to the host vehicle 10 and the speed of the host
vehicle 10. Additionally, the processor 50 may further process the
size and shape of the object in order to determine whether the
object will likely collide with the host vehicle 10 and, whether
the object is of a sufficient size to be a concern upon impact with
the host vehicle 10. If the object is determined to be sufficiently
small in size or moving at a sufficiently slow rate, the object may
be disregarded as a potential crash threat, whereas a large object
moving at a sufficiently high rate of speed toward the host vehicle
10 may be considered a crash threat.
[0049] While the crash sensing system 20 is described herein in
connection with an integrated IR transmitter/receiver having nine
IR transmitters and nine photosensitive receivers each arranged in
an array of three-by-three (3.times.3), and the addition of an
additional IR transmitter 26 and an optional photosensitive
receiver 27, it should be appreciated that other infrared transmit
and receive configurations may be employed without departing from
the spirit of the present invention. It should further be
appreciated that other shapes and sizes of coverage zones for
transmitting IR radiation and receiving photosensitive energy
radiation may be employed and that the transmitters and/or
receivers may be located at various locations on board the host
vehicle 10. U.S. Patent Application Publication No. 2009/0099736,
entitled "VEHICLE PRE-IMPACT SENSING SYSTEM AND METHOD" discloses
various configurations of IR transmitter and receiver arrays for
detecting objects to a lateral side of a vehicle 10. The
aforementioned U.S. Patent Application Publication is hereby
incorporated herein by reference. It should be further be
appreciated that variations in segmented lens or reflector designs
may be utilized to provide design flexibility for customized
coverage zones. One example of a segmented lens is disclosed in
U.S. Patent Application Publication No. 2008/0245952, filed on Apr.
3, 2007 and entitled "SYNCHRONOUS IMAGING USING SEGMENTED
ILLUMINATION," the entire disclosure of which is hereby
incorporated herein by reference.
[0050] It should be appreciated that a complete field image
encompassing all the coverage zones may be generated every scan of
the entire array of the covered volume. By comparing successively
acquired photosensed images, the size, shape, location, range and
trajectory of an incoming object can be determined. To aid in the
estimation of the range (distance) of the object from the system
20, and hence the host vehicle 10, the additional IR illuminator 26
and optional receiver 27 may be employed along with triangulation.
By employing triangulation, the presence of an object in the
designated zones is compared to the additional IR transmit zone 28
such that the range (distance) can be determined. Additionally, the
reflection power of the signal received can be used to enhance the
range estimate and thereby enhance the detection of a pre-crash
event.
[0051] The vehicle pre-crash impact sensing system 20 employs a
feature detection scheme that identifies certain features of an
object vehicle, particularly an object vehicle moving laterally
toward the host vehicle 10, to provide enhanced vehicle
discrimination. According to a first embodiment of the feature
identification scheme, the sensing system 20 employs the
horizontally illuminated IR calibration beam 28 in conjunction with
the IR transmit beams 32A-32I in an attempt to identify a known
feature such as bumper and/or grille of a laterally oncoming
vehicle. In doing so, the horizontal calibration IR beam
transmitter 26 is multiplexed between the main IR beams 32A-32I to
allow enhanced range calibration. By multiplexing the additional IR
illumination calibration beam 28 with the standard nine IR transmit
zones 32A-32I, the range of the reflected object can be better
estimated. Additionally, by employing a calibration chart, the
reflection coefficient of the surface of the object detected may be
used for increased accuracy range estimate, and thus improved risk
assessment for proper side air bag deployment or other
countermeasure. The optional tenth photosensitive receiver 27 may
be employed to provide received photosensitive signals within zone
29 to further enhance the process.
[0052] According to another embodiment, enhanced oncoming laterally
moving vehicle discrimination can be achieved by employing one or
more scanned beams in a generally U-shape configuration which is
generally configured to encompass the shape of a common vehicle
front, particularly the fascia. Multiple U-shaped patterns
extending from a distant focus to larger and nearby may be created
with the physical structure of the beam hardware (e.g., via the
optical design). Alternately, the beam pattern can be created in
software if the number of beams fully covers the oncoming laterally
moving vehicle's trajectory path from far to nearby. The U-shaped
beam forms may have ends of about three feet by three feet which
focuses on the oncoming laterally moving vehicle's headlamps/signal
markers and a center connecting line of about a two foot height
which receives the oncoming laterally moving vehicle's grille
chrome. Accordingly, the pre-impact sensing system applies vehicle
fascia detection with vehicle front grille shaped optical regions
for improved detection of approaching vehicles. Nine overlapping
regions may allow target tracking and relative ranging, and the
geometry can apply to either the IR illumination or light receiver
shape or possibly both the transmit and receiver shapes.
[0053] According to a further embodiment, enhanced oncoming
laterally moving vehicle discrimination may be achieved by
detecting the differential spectral return of highly reflective
vehicle surfaces, such as signal markers, headlamps, fog lamps,
license plates, chrome and other features typical of vehicle front
ends. Additionally, background light illumination levels may also
be used to measure the highly reflective vehicle elements.
Additionally, the system 20 may be used to detect the headlamp on
status of the oncoming vehicle further, thereby allowing
discrimination of their presence as well as any possible pulse
width modulation (PWM). LED headlamps which are also pulsed may be
sensed and used as an additional discrimination element. The
geometry of the spectral objects on an inbound oncoming laterally
moving vehicle may also aid in the discrimination risk
assessment.
[0054] Referring now to FIG. 9, a routine 200 is illustrated for
sensing a pre-crash event of a vehicle to deploy one or more
devices and includes steps of implementing the various
aforementioned embodiments of the vehicle feature identification to
advantageously provide updated object range estimates. Routine 200
begins at step 202 and proceeds to step 204 to illuminate the IR
beam N, wherein N is the number of IR transmit beams which are
sequentially activated. Next, in step 206, routine 200 stores the
IR transmitter that is turned on, and receives the IR amplitude
data for the N.sup.th receive zone. It should be appreciated that
the transmit beams are turned on one beam at a time, according to
one embodiment, and that the light energy including IR energy
reflected from objects within each zone is detected by receivers
covering the corresponding receive zones. In step 208, beam N is
turned off as part of this process, and at step 210, the IR off is
stored and the amplitude data of received light energy for beam N
is stored with the IR off. Next, N is incremented to the next
successive zone of coverage. At decision step 214, routine 200
determines if N has reached a value of ten which is indicative of
the number of IR transmitters and, if not, repeats at step 202.
Once a complete cycle of all ten zones has been completed with N
equal to ten (10), routine 200 proceeds to step 216 to indicate
that a frame is complete, such that the received energy data is
mapped out for the coverage zones.
[0055] Once a frame is complete, routine 200 proceeds to step 218
to perform a sequential IR modulation signal calculation for each
of the IR sensors, shown as sensors 1-9 indicative of the outputs
of the respective photodetectors 24A-24I. Step 218 essentially
performs a sequential IR modulation signal calculation by taking
the difference of the sensed photo signal while the infrared
transmitter is turned on and then when the infrared transmitter is
turned off for each coverage zone. As such, for zone 1, a signal
(X1) is determined based on the difference of the receive signal
with the IR on and the IR off, signal (X2) is indicative of the
receive signal with the IR on minus the IR off, etc. By turning the
IR transmitter array on and off at a frequency such as three
hundred (300) hertz, the emitted IR beams are essentially modulated
at the switching or modulation frequency. The difference between
the received IR energy signals when the IR transmitter is turned on
and when the same IR transmitter is turned off produces a ripple
signal as described herein. The sequential IR modulation signal
calculation is performed for each of the nine zones 1-9 to generate
corresponding ripple signals to remove or get rid of the overall
background noise.
[0056] For an embodiment that employs the tenth IR receiver 27,
routine 200 performs step 220 which processes the output of the
tenth IR receiver 27 to acquire enhanced signal to noise ratio.
Step 220 is an optional step that performs a reference IR
modulation signal calculation for the tenth receiving sensor, also
referred to as receiver 27. In doing so, signal (X10) is generated
as a function of the IR on minus the IR off for the tenth coverage
zone.
[0057] Routine 200 then proceeds to step 222 to perform a
differential signal calculation (Y) for each of sensors 1-9 to
acquire an enhanced differential signal by eliminating or removing
background noise. The differential signal calculation involves
calculating the difference between signal X1 and signal X10, to the
extent a tenth transmitter is employed. Similarly, the signal Y2 is
acquired by taking the difference between signal X2 and signal X10.
Similarly, each of zones 3-9 involves subtracting the signal X10
from the corresponding received signal for that zone to provide
respective ripple signals for each zone.
[0058] Following step 222, routine 200 proceeds to step 224 to use
an object calibration for vehicle feature identification. According
to one embodiment, routine 200 uses a standard object calibration
for a bumper detection to detect the bumper or similar feature(s)
on a lateral oncoming vehicle. According to another embodiment,
routine 200 employs a highly reflective object calibration for
fascia and headlamp detection. It should be appreciated that the
object calibration for bumper detection and reflective object
calibration for fascia/headlamp detection may be achieved by use of
one or more calibration charts or look-up tables, such as the
exemplary calibration chart shown in FIG. 10.
[0059] The calibration chart shown in FIG. 10 essentially maps a
plurality of different sample objects having various features such
as shapes, colors, materials (textures) and reflectivity as a
function of the ripple signal in hertz versus range in feet. It
should be appreciated that the IR receiver photodetectors each
provide an output frequency dependent upon the photosensitive
detection. The frequency is generally equal to the amplitude (A) of
the ripple signal divided by the distance (d) squared
(frequency=A/d.sup.2). The ripple signals shown are the reduced
noise Y signals generated at step 222. The various targets that are
shown by lines or curves 70A-70H in the example calibration table
of FIG. 10 include arbitrary selected materials such as a white
paper, black cotton twill, brick, tree and fog, black paper, black
board, black cotton gauze, asphalt, and a typical sky scenario. It
should be appreciated that these and other targets or selected
materials may be mapped out in a given calibration table for use
with routine 200.
[0060] Next, in step 226, routine 200 compares the differential
signal calculations Y to determine the highest correlation for a
detected geometry. In doing so, routine 200 uses a selected
calibration map to determine the optimum range estimate based on a
common range from the nine received ripple signals. In step 228,
routine 200 updates the object range estimate based upon the
identified feature(s). Accordingly, it should be appreciated that
by identifying an anticipated feature for the front end of a
lateral approaching vehicle, such as the vehicle bumper, fascia
and/or headlamp, enhanced object range may be estimated for use in
the pre-crash sensing system 20.
[0061] Following the step of updating the object range estimate,
routine 200 proceeds to decision step 230 to determine if the
temporal gating has been met, and if not, returns to step 202. The
temporal gating step 230 may be the same or similar temporal gating
step described in connection with step 108 of routine 100. If the
temporal gating has been met, routine 200 proceeds to decision step
232 to see if the thermal IR safing has been enabled and, if so,
deploys one or more devices in step 234. If the thermal IR safing
is not enabled, routine 200 returns to step 202. The thermal safing
step 232 may be the same or similar to the safing logic described
in connection with step 110 of routine 100.
[0062] Accordingly, routine 200 advantageously provides for an
enhanced object range estimate based on the detected type of
feature of a vehicle that is oncoming in the lateral direction. By
detecting one or more features of the detected object, routine 200
advantageously looks up the calibration data and provides an
updated range estimate which advantageously enhances the
determination of whether a laterally approaching vehicle, such as
an automobile, is expected to collide with the host vehicle 10.
[0063] Referring to FIG. 11, the U-shaped geometry for a single
beam geometry transmit beam 32 is shown superimposed onto the front
side fascia portion of the front side of a vehicle, which
represents a potential oncoming laterally moving vehicle. The
U-shaped transmit beam 32 may be broadcast as an infrared beam by a
single IR transmitter, according to one embodiment. According to
one embodiment, multiple U-shaped IR transmit beams 32A-32I are
transmit as shown in FIG. 12, each having a generally U-shape and
covering separate zones, which overlap each other. The multiple
U-shape beams may extend from a distant focus to larger and nearby
and could be created with the physical structure of the beam
hardware (e.g., via the optical design). Alternately, such a beam
pattern could be created in software if the number of beams fully
covers the oncoming laterally moving vehicle's trajectory path from
a far distance to a nearby distance. According to one example, the
U-shaped beam form may have ends of about three feet by three feet
at distance of about six feet, which focuses on the oncoming
laterally moving vehicle's headlamp/signal markers and a center
connecting line of about a two foot height which receives the
oncoming laterally moving vehicle's grille chrome. While an array
of U-shape transmit beams is shown and described herein, it should
be appreciated that the geometry of the light receivers may be
shaped in a U-shape instead or in addition to the transmit
beams.
[0064] By processing the U-shaped beam and the IR signals received
therefrom, the trajectory and range of the oncoming vehicle object
may be better determined. In an alternate embodiment, the U-shaped
beams may be shaped in an oval shape for simplicity, or another
shape that picks up the fascia and similar features of the front
end of the oncoming vehicle. It should be appreciated that as the
IR transmit beams cross horizontally the front fascia of an
oncoming front end of a vehicle, a resultant ripple signal is
generated. The ripple signal is the difference in detected energy
signal when the IR is turned on and when the IR is turned off. It
should be appreciated that for a high reflectance feature, such as
the headlamps and signal markers, a higher ripple signal is
achieved having a higher frequency. Typically, the center grille
area of a front end of an oncoming vehicle is mainly paint which
has a lower reflective coefficient versus the reflected components
of the headlamps and the signal markers. The ripple signal
signature can be processed to determine the presence of a likely
vehicle fascia or portion thereof including the headlamps and
signal markers of the front end of an oncoming vehicle.
[0065] The pre-crash sensing system 10 further employs a modulation
technique to nullify background ambient light conditions and better
enhance the estimated target range. The extreme lighting variation
from darkness to full sunlight presents many challenges to an
object detection system. These and other deficiencies can be
overcome by turning the IR transmit array on and off at a high
frequency of three hundred (300) hertz, for example, or otherwise
provide amplitude modulation of the IR light source with a square
wave or a sine wave so as to nullify the background ambient light
conditions and better enhance the estimate of target range.
[0066] Extreme lighting variations and other deficiencies may be
overcome by a modulation technique which measures a scene with
ambient lighting and also with an artificial IR illumination
source. The difference between the two measurements provides an
inferred target range within the scene. This modulation method
provides a very cost-effective method of target ranging, yet does
not require extreme power levels as created by typical solar
exposure which can be cost prohibitive. According to another
embodiment, the modulation technique may be implemented with a
carrier signal as disclosed in U.S. Pat. No. 6,298,311, entitled
"INFRARED OCCUPANT POSITION AND RECOGNITION SYSTEM FOR A MOTOR
VEHICLE," the entire disclosure of which is hereby incorporated
herein by reference.
[0067] Referring to FIG. 13, the modulation technique is
illustrated in connection with an example of sensed IR and ripple
signal both when a parked vehicle is located in the lateral side
detection zone of the host vehicle 10 and when the parked vehicle
is removed from the detection zone. As shown, the sensed raw
irradiance or received light energy level is plotted as a function
of time versus frequency as the host vehicle drives by the parked
car shown at top at time of about eighteen (18) seconds to a time
of about thirty-three (33) seconds, such that initially there is no
car on the lateral side of the host vehicle up until time equals
eighteen (18) seconds, and the car passes through the detection
zones and then departs the detection zones at time equals
thirty-three (33) seconds. Initially, sunlight reflected from
asphalt is detected at line 90 and a shadow is detected as
indicated. Once the parked vehicle enters the detection zone, the
received raw IR irradiance or light energy indicated by area 92
between lines 94 and 96 increases between max and min values when
the IR transmitter is turned on at max border 94 and when the IR
transmitter is turned off at min border 96. The difference between
the IR turned on and off signals is represented by the ripple
signal 98 which goes from approximately zero (0) frequency to a
frequency of about five thousand (5,000) when the lateral location
vehicle passes through the detection zone, and returns to a value
of near zero (0) when the lateral located parked vehicle departs
the detection zone. The modulation techniques advantageously allows
for detection of an object due to reflected signal independent of
shadow or sunlight.
[0068] The data shown in FIG. 13 illustrates frequency data where
the object was generally in the center of the images and the IR
illumination alternates from on to off. This data yielded a ripple
signal 98 of around one hundred (100) hertz of variation in the raw
data when looking at the asphalt, and about five thousand (5,000)
hertz with a white target at about six (6) feet, according to one
example. The ripple signal 98 did not change even though the
ambient lighting moved from a full sun exposure from the time
period of about zero to ten (0 to 10) seconds to the shadow of the
vehicle at about the time period of about seventeen (17) or
eighteen (18) seconds. The data shown in FIG. 13 was taken within a
950 nanometer band which requires rather expensive optical bandpass
filters to eliminate sunlight. The modulation technique allows for
operation with a much less costly high wavelength filter, such as
greater than 700 nanometers.
[0069] Given the ripple signal 98 generated in the table of FIG.
13, the modulation technique enables the range estimation to be
enhanced with acquisition from the calibration table shown in FIG.
10. As seen in FIG. 10, a white object's range can be fairly well
predicted as shown by the curve 70A representing white paper. Dark
objects are also predictable as shown by the curve 70H representing
asphalt. Textured fabrics may exhibit a self-shadowing effect. By
use of this inferred range information and the event progression of
the multiple spot data, a side impact risk estimate can be made
and, if deemed of sufficient risk, the side air bags can be
deployed. By monitoring the ripple current 98 shown in FIG. 13, for
each coverage zone, the ranges as shown by lines 72A, 72B and 72C
may be established, for example. By looking at common points within
the set of ranges 72A-72C, an enhanced range estimate may be
established. In the given example, ranges 72A-72C have a common
range value of about six feet. Since the frequency output of the IR
photodetectors may vary based upon color or reflectivity in
combination with a range, an enhanced range estimate may be
provided by looking for common range values within the zones.
[0070] The pre-crash sensing system 10 further employs a terrain
normalization method to normalize out stationary objects that pass
through the detection zones to the lateral side of the vehicle 10.
Referring to FIG. 14, an example driving scenario is shown
illustrating a host vehicle 10 employing the sensing system 20 is
traveling along a roadway relative to several objects including an
angled barrier 80, a lateral oncoming vehicle 84, a passing lateral
vehicle 82, and a laterally projecting vehicle 86 expected to
collide with the host vehicle 10. The terrain normalization method
is able to normalize out stationary objects, such as the angled
barrier 80 which, due to its shape, may appear to be moving toward
the host vehicle 10, such that the system 20 may ignore the
stationary object 84 in determining whether or not an impending
lateral collision will occur. By use of range data, which is
inherent to the power level of the reflected signal, a
three-dimensional volume can be estimated with the array of
coverage zones 34A-34I. As the host vehicle 10 is driven forward,
the frontmost beams will detect a terrain pass by of an object.
This terrain information, including inferred distance, is
propagated to the successive rearward beams which follow and is
used to normalize out the existence of stationary objects which are
characterized as clutter to be ignored. Objects which have a
lateral velocity component are recognized as potential oncoming
targets and their characteristics are further evaluated for
assessed threat of a lateral collision to the host vehicle 10.
[0071] In particular, physical objects such as an angled barrier
(e.g., a guardrail) or an angled road line may have a shape
resembling an oncoming vehicle bumper and can produce similar or
identical IR pattern signatures which possibly could cause
undesirable deployment of an air bag or other device when not
properly detected. The terrain normalization method attempts to
detect and eliminate such false triggers. With the transmitter and
receiver arrangement shown, a matrix of infrared beams and
receiving beams illuminate the side of the host vehicle 10 to
provide a sensed volume by the three-by-three (3.times.3) array.
Using range data, which is inherent to the power level of the
reflected signal, a three-dimensional volume can be estimated. As
the host vehicle 10 is driven forward, the frontmost beams or zones
will see the terrain pass by first. This terrain information
includes distance that is propagated to subsequent following beams
which follow behind in progression and are used to normalize out
stationary objects. Objects which have a lateral velocity component
are recognized as potential oncoming targets to the host vehicle
10, and their characteristics are further evaluated for assessed
threat to the host vehicle 10.
[0072] Each scan of the matrix yields an object light level for
each spot of the zones detected. From this, the range (distance) of
an object can be inferred according to a distance look-up table.
Generally, objects which are at ground level are quite low in
reflected energy as power is related to one divided by the distance
squared. As seen in FIG. 14, barriers and oncoming cars illuminated
by using the leading spot zones and propagating this information
rearward to the trailing spots is achieved with a normalization
technique. The data may be processed by averaging, normalization,
and time differential, amongst other embodiments. Additionally,
road speed and steering angle could be used to propagate the
optical distance of the measured lead objects rearward, thus,
eliminating them from causing false deployments. In a similar
manner, the detection of objects entering the rearmost zones first
are processed in reverse, such as when a car passes the host
vehicle 10. The resultant matrix of processed data which has been
normalized to remove oncoming and passing objects is used to
evaluate the event propagation of a lateral object, such as an
oncoming laterally moving vehicle 86.
[0073] Referring to FIGS. 15A-15D, the progression of a stationary
object is illustrated as a host vehicle 10 passes by stationary
object 84. As seen in FIG. 15A, the stationary object 84 is in
front of the nine detection zones. As the host vehicle 10 moves
forward to the lateral side of a stationary object 84, the
stationary object 84 is shown first being detected by zone A1 in
FIG. 15B, and then detected by both zones A1 and A2 in FIG. 15C,
and then detected by zones A1-A3 in FIG. 15D. Finally, as the
object 84 departs at least part of the detection zone, the object
84 is shown in FIG. 15E departing zone A1 and is still detected in
zones A2 and A3. In this scenario, the system 20 employs a
normalization routine to subtract out the propagated signal
detected in a forward located zone from the raw data in an attempt
to detect whether or not a lateral motion of the object is
occurring. As seen in FIGS. 15A-A-15E-E, the detection of the
object 84 in zone A1 is detected and as it approaches zone A2, the
propagated signal of zone A1 is subtracted from the raw data of
zone A2 to provide a normalized result indicative of a stationary
object which should be rejected.
[0074] Referring to FIGS. 16A-16D, a scenario is shown in which a
stationary object 80 in the form of an angled barrier, for example,
is passed by the host vehicle 10 and the system 20 employs terrain
normalization to reject the stationary object 80, despite its
potentially deceiving shape due to its angle towards the vehicle
10. In this example, the stationary object 80 first enters zone A1
in FIG. 16A, and then proceeds to enter zone A2 in FIG. 16B. In
FIG. 16C, the stationary object 80 is detected by zones A1, A2 and
B1, and then in FIG. 16D, the stationary object 80 is detected
primarily in zones A3, B2 and C1. As the stationary object passes
from zone A1 through and into zones A2, A3 and B1, the terrain
normalization subtracts the propagated signal of the leading zone
from the following zone. For example, in zone A2, the normalized
result is that the propagated signal of zone A1 is subtracted from
the raw data of zone A2 to determine that the stationary object 80
is detected and should be rejected. The propagated signal is
subtracted based on a delay time which is determined based on the
speed of host vehicle 10, such that the signal data of the
preceding zone captures an area of space that is also captured in
the following zone. The terrain normalization thereby takes into
consideration the detected signal information from the preceding
zone taking into consideration the time delay and the speed of the
host vehicle 10.
[0075] Referring to FIGS. 17A-17D, a further example of a laterally
oncoming vehicle 86 is illustrated in which the laterally oncoming
vehicle 86 first enters zone A1 and A2 in FIG. 17B and proceeds
into zones B1, B2, A1, A2 and C1 in FIG. 17C, and finally proceeds
into zones A1, A2, B1, B2 and C1-C3 in FIG. 17D. The terrain
normalization effectively removes stationary objects so that the
system 20 can detect that the laterally oncoming vehicle 86 is not
stationary, but instead is moving with a lateral velocity component
toward the host vehicle 10, such that an oncoming collision may
occur.
[0076] Referring to FIGS. 18A-18C, the terrain normalization
routine 300 is illustrated, according to one embodiment. The
terrain normalization routine 300 begins at step 302 and proceeds
to step 304 to illuminate the outermost row of IR transmit beams in
zones A1, A2 and A3. Next, in step 306, routine 300 receives the IR
amplitude data for receive zones A1, A2 and A3 while the IR is
turned on. Next, the IR transmit beams for zones A1, A2 and A3 are
turned off. Proceeding to step 310, routine 300 receives the IR
amplitude data for zones A1, A2 and A3 while the IR transmit beams
are turned off and stores the received amplitude data while the IR
transmit beams for zones A1, A2 and A3 are turned off. At step 312,
routine 300 turns on the IR transmit beams for the next or middle
row of zones B1, B2 and B3. In step 314, routine 300 receives the
IR amplitude data for zones B1, B2 and B3 while the IR transmit
beams are turned on and stores the received IR amplitude data in
memory. At step 316, routine 300 turns off the IR transmit beams
for zones B1, B2 and B3. With the IR transmit beams turned off,
routine 300 proceeds to step 318 to receive the IR amplitude data
for zones B1, B2 and B3 and stores the received IR amplitude data
in memory.
[0077] Routine 300 proceeds to step 320 to turn on the IR transmit
beams for the third or closest row of zones C1, C2 and C3. Next, at
step 322, routine 300 receives the IR amplitude data for zones C1,
C2 and C3 and stores the received IR amplitude data in memory. In
step 324, routine 300 turns off the IR transmit beams for zones C1,
C2 and C3. With the IR transmit beams turned off, routine 300
proceeds to step 326 to receive the IR amplitude data for zones C1,
C2 and C3 and stores the received IR amplitude data in memory.
Next, routine 300 turns on the IR transmit beam for the tenth or
additional transmit beam at step 328. With the tenth or additional
IR transmit beam turned on, routine 300 proceeds to step 330 to
receive the IR amplitude data for zones A1, A2, A3, B1, B2, B3, C1,
C2, C3, and the tenth receiver while the tenth or additional IR
transmit beam is turned on. Finally, the initial frame is complete
at step 332.
[0078] Once the frame is complete, routine 300 proceeds to step 334
to perform a sequential IR modulation signal calculation X for each
of sensors one through nine, shown labeled A1-C3. This includes
calculating the difference in signals when the IR transmit beam is
turned on and when the IR transmit beam is turned off for each of
zones A1-C3 to provide a raw ripple signal for each of zones
A1-C3.
[0079] Proceeding on to step 336, routine 300 stores a history of
the IR modulation signal calculation (X) for each of sensors one
through nine for zones A1-C3. This involves storing the time
average of the IR signals for each of the zones A1-C3.
[0080] At step 338, routine 300 looks at the forward vehicle motion
and cancels out the stationary signals from the following sensor
locations. This includes normalizing the signal for a given zone by
subtracting out from the following zone the history of the previous
zone. For example, the history of zone A1 is subtracted from zone
A2 when an object passes from zone A1 to zone A2. The continued
signal normalization applies to zone A3 in which the history from
zones A1 and A2 is subtracted from zone A3 at the appropriate time
based upon a time delay based on the vehicle speed. The time delay
is based on vehicle speed so that the zones cover the same area of
space. Signal normalization also occurs in rows B and C by
subtracting out the signal from the prior zone. It should be
appreciated that the same signal normalization applies in the
reverse direction for a vehicle or object passing laterally from
the rear of the host vehicle 10 toward the front of the host
vehicle 10, except the signal normalization is reversed such that
the propagating signal in zone A3 is subtracted from A2, etc. For
example, for zone A2, you take the current X2 data and subtract off
the history of zone A3 such that a sliding window essentially is
provided. The terrain normalization essentially eliminates the fore
and aft movement parallel to the host vehicle 10 in the detection
zones. By doing so, stationary objects or clutter are rejected.
[0081] Routine 300 then proceeds to step 340 to determine the
lateral component of the moving object. The lateral component of an
object is based on the lateral movement toward or away from the
lateral side of the host vehicle 10. Next, in decision step 342,
routine 300 determines if there is lateral velocity component
greater than eighteen miles per hour (18 mph) and if the object is
large enough and, if so, proceeds to step 346 to update the object
range estimate, and then proceeds to decision step 348 to check the
temporal gating. If the object is not large enough or if the
lateral velocity component is not greater than eighteen miles per
hour (18 mph), routine 300 returns to step 302. At the decision
step 348, temporal gating is compared to determine whether or not
an object is likely to collide with the host vehicle 10 and, if so,
routine 300 proceeds to decision step 350 to determine if thermal
IR safing is enabled and, if so, deploys an output at step 352. If
the thermal IR safing is not enabled, routine 300 returns to step
302. It should be appreciated that the temporal gating of step 348
and the thermal IR safing of step 350 may be the same or similar to
those steps provided in routine 100 as discussed above.
[0082] While a thermal far IR safing receiver 46 is shown and
described herein for providing thermal IR safing, it should be
appreciated that other safing techniques may be employed to
eliminate false triggers. As described, a matrix of IR beams
illuminates the side of host vehicle 10 to provide a sensed volume.
A single IR beam may be provided in a matrix of beams. Using range
data, which is inherent to the power level of the reflected signal,
a three-dimensional volume can be estimated. The addition of a
separate technology to "safe" the primary deploy signal is required
to ensure against false air bag deployment. "Safing" is defined as
a complementary measure to verify that the detected object is an
oncoming laterally moving vehicle where the measured speed of the
oncoming laterally moving vehicle matches the speed measured by the
primary detection. Moreover, the measured speed may be
approximately fifteen (15) to fifty (50) miles per hour, according
to one example. Additionally, this concept of lateral velocity
verification can be used to enable sub-fifteen mile per hour air
bag deployment.
[0083] Use of discrimination sensors to assess data to do a lateral
velocity calculation of the oncoming laterally moving vehicle
compared to the safing lateral velocity may be provided. If the two
are similar, it can be assumed with a high degree of confidence
that the oncoming object is indeed a sufficient threat to the
occupants of the host vehicle 10. Objects which have a significant
lateral velocity component, such as those greater than eighteen
(18) miles per hour, may be recognized as potentially dangerous
targets and their characteristics may be evaluated for assessed
threat to the host vehicle 10. Each scan of the matrix yields an
object light level for each spot or zone. An analysis of the light
levels from all the spots can infer the distance, velocity and
trajectory of an object from the host vehicle 10.
[0084] Discrimination technology considered for the safing
technique can include active near IR (NIR) radar, or camera. One or
more safing technology and one or more deploy technology may be
utilized in the design. Individual safing or deploy technologies
can include active near IR, far IR (FIR), ultrasonic acoustics,
laser time-of-flight sensor (3D scanner), 3D camera, or stereo
camera.
[0085] Employing the thermal IR safing technique, heat may be
detected from an oncoming vehicle. According to an ultrasonic
sensed safing technique, ultrasonic sensors perform the safing
function. The safing function is employed to ensure the event
progression is due to an incoming vehicle and not due to other
situations that would not require a side air bag deployment. Safing
may prevent deployment of a side air bag when the host vehicle 10
strikes a stationary object, such as a tree, pole or parked
vehicle. These objects are not likely to have the thermal signature
of a moving vehicle, and in extreme yaw conditions, may not return
a steady ultrasonic return which is especially true for trees and
poles. Therefore, there is a need to relate a means to disable or
reduce safing requirements in yaw conditions.
[0086] Situationally dependent safing is a method to modify side
air bag pre-impact deployment safing based upon the vehicle
stability. During normal vehicle conditions, an active IR sensing
system is employed to determine when a side impact is imminent.
Supplemental information from either an ultrasonic or passive IR
system is used for safing. If the host vehicle 10 path is tangent
to its four-aft direction or the target follows a linear path into
the side of the host vehicle 10, incoming targets will follow a
normal progression and safing techniques will provide information
necessary to make a reliable decision. A relatively linear
progression will allow sufficient path information of the active IR
system to generate a mature path. In extreme yaw conditions,
however, the path of the host vehicle 10 may not allow the
development of a mature track for impacts. Moreover, the
supplemental safing sensors are less likely to provide adequate
information to supplement the deployment condition. If an
ultrasonic sensor is employed, the host vehicle 10 may spin into
the target too quickly to provide an adequate return. If a passive
IR system is employed, the target may not generate the thermal
signature necessary to allow deployment. Therefore, a decision tree
may allow for safing levels to be reduced in cases where the host
vehicle 10 is not following a consistent path due to the high yaw.
The decision tree may include logically ORing the following
requirements: Far IR (FIR) is greater than threshold, steering
wheel angular rate is greater than N degrees per second, yaw is
greater than N degrees per second, external lateral slip divided by
yaw, and lateral acceleration greater than 0.5 g. The output of the
OR logic is then logically ANDed with the discrimination output to
determine whether or not to deploy one or more devices.
[0087] Additionally, vehicle travel direction can be inferred by
the ground terrain monitoring of the active IR scanning system. By
use of both left and right pre-crash side sensors to monitor the
optical flow of asphalt pattern directions, the vehicle velocity
and direction including lateral sliding can be detected and
approach countermeasures initiated. Under inclement conditions,
such as blowing snow/rain/sand side slip monitoring is failed-safed
by a lateral yaw rate sensor. Lateral sliding information can be
used for side air bag threshold lowering, stability control, or
potentially rollover detection. The lateral slip sensor may use a
left or right sensor to monitor road pattern directions. With the
transmit/receive beams, an optical flow through beam matrix
determines ground travel direction, vehicle rotation and potential
vehicle roll.
[0088] As mentioned herein, the array of transmit and receive IR
beams may be arranged in an overlapping configuration. Tailoring of
three-dimensional volume to the side of the host vehicle 10 can
pose a challenge to ensure an oncoming laterally moving vehicle is
detected and a side air bag is deployed, yet allow the numerous
no-deploy objects which pass harmlessly by the host vehicle 10 to
not cause false deploys. The beam overlap may allow increased
spatial resolution with a minimum number of discrete beams by use
of the beam overlap. In order to increase the effective spots or
zones of a pre-crash side impact sensor without adding more
channels, each of the nine beam spots may be enlarged to allow a
twenty percent (20%) beam spot overlap which provides twenty-one
multiplexed zones, in contrast to the above disclosed nine
non-overlap zones, which allows increased distance resolution of
the incoming object. The geometry can apply to IR illumination or
light receiver shape or possibly both transmitter and receiver
shapes for more resolution. Accordingly, the beam overlap method
may consist of overlapping scanned areas or regions to allow
increased target tracking resolution.
[0089] Accordingly, the pre-crash sensing system 20 of the present
invention advantageously detects an impending collision of an
object with the host vehicle 10, prior to the actual collision. The
crash sensing system 20 is cost affordable and effective to
determine whether an object is approaching the host vehicle 10 and
may collide with the vehicle, sufficient to enable a determination
of the impending collision prior to actual impact. Further, the
pre-crash sensing system 20 may determine whether the object is of
sufficient size and speed to deploy certain countermeasures.
[0090] It will be understood by those who practice the invention
and those skilled in the art, that various modifications and
improvements may be made to the invention without departing from
the spirit of the disclosed concept. The scope of protection
afforded is to be determined by the claims and by the breadth of
interpretation allowed by law.
* * * * *