Data processing: vehicles – navigation – and relative location – Vehicle control – guidance – operation – or indication – Vehicle subsystem or accessory control
Reexamination Certificate
2002-01-17
2003-12-09
Cuchlinski, Jr., William A. (Department: 3661)
Data processing: vehicles, navigation, and relative location
Vehicle control, guidance, operation, or indication
Vehicle subsystem or accessory control
C701S047000, C180S268000, C180S271000, C382S103000, C382S104000, C382S106000, C280S735000, C340S438000
Reexamination Certificate
active
06662093
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates in general to systems or methods used to determine when an airbag should be deployed to protect an occupant in a vehicle from the impact of a collision. In particular, the present invention relates to inferring three-dimensional occupant characteristics from a two-dimensional segmented image of the occupant, and determining when an airbag should be deployed using those occupant characteristics in an iterative and probability-weighted manner to determine if a vehicle collision has occurred.
Conventional collision detection systems typically use accelerometers or weight-based sensors to determine if there has been a vehicle collision to trigger the deployment of an airbag. Such systems are subject to false alarms from severe road conditions, such as a vehicle sliding on ice the bumps into a curb, and minor impacts, such as the accidental hitting of a parking block while entering a parking lot. It would be desirable for an airbag deployment system to be based on occupant characteristics derived from an image of the occupant because the image of an occupant is less susceptible to errors caused by rapidly shifting movement or weight.
Even prior art systems that do not rely solely on weight-based determinations are susceptible to errors due to “noise” because such systems focus solely on the most recent measurement or sensor input, and ignore the series of measurements or inputs captured mere fractions of a second earlier. Measurement “noise” results from several factors, including the inherent imperfections of capturing sensor inputs. It would be desirable for an image processing system or an airbag deployment system to utilize an iterative process that would integrate the information contained in the most recent image into a comprehensive framework that includes prior predictions and indirectly, the prior images used to make those prior predictions. It may also be desirable for some or all predictions to be probability-weighted predictions and for such weighted predictions to incorporate probabilities associated with predefined occupant states such as leaning left towards the driver, leaning right away from the driver, or sitting upright. It may also be desirable for a system to incorporate predefined occupant modes such as crashing, pre-crash braking, stationary (asleep), or human (normal).
Existing camera-based and image-based systems are typically limited by the fact that they rely on two-dimensional sensor readings. The images captured by cameras, including video cameras, are inherently two-dimensional images. However, some important characteristics such as volume (which can be used to calculate mass) are three-dimensional. It would be beneficial if three-dimensional information could be inferred from a series of two-dimensional images taken by a single fixed video camera. Moreover, it would be helpful if predefined occupant states and modes were incorporated into the iterative process of deriving a three-dimensional information from a series of two-dimensional images.
Conventional airbag deployment systems have contributed significantly to the safety of occupants in automobile crashes. However, there may be occasions when deployment of an airbag is not desirable in the context of a vehicle collision. A collision of minimal impact may not merit the deployment of an airbag, especially if the occupant is wearing a seat belt. Moreover, the close proximity of an occupant to the deploying airbag may make airbag deployment undesirable in a particular situation. It would be desirable for an airbag deployment system to track and predict the location of an occupant so that the airbag deployment system can be disabled in situations where the occupant would be too close to the airbag deployment system.
Conventional airbag deployment systems tend to take an all or nothing approach with respect to airbag deployment. Under such approaches, an airbag is either fully disabled or is deployed at full strength. A total disablement of the airbag precludes the occupant from receiving any benefits of an airbag, a generally useful safety device. A full strength deployment of an airbag may subject the occupant to undesirable force in a low speed crash. It would be desirable if an airbag deployment system could be deployed at various different strengths depending on the impact of the occupant and the airbag. It would be desirable if metrics such as kinetic energy, momentum, or other measurements utilizing the characteristics of mass and velocity were used to determine the magnitude of impact that an airbag needs to absorb from the impacting occupant. As discussed above, current methods for determining occupant characteristics such as mass, velocity, location, and other characteristics suffer from significant limitations in the existing art.
SUMMARY OF THE INVENTION
This invention relates to an image processing system used to determine whether a vehicle has been subjected to a collision requiring the deployment of an airbag. The invention uses a tracking and predicting subsystem to incorporate past sensor readings and predictions along with a current sensor reading to generate one or more occupant characteristics relevant to a crash determination. In a preferred embodiment, the system uses mathematical heuristics such as a Kalman filter to track and predict occupant characteristics, and such heuristics can be applied to interacting multiple models in a probability-weighted fashion. A crash detection subsystem processes one or more occupant characteristics from the tracking and predicting subsystem to determine whether a crash has occurred. A preferred method of crash detection is for the crash detection subsystem to compare an occupant characteristic with an appropriate metric relating to one of the various models, and determine which of the models (which includes crash model) best describes the occupant characteristic.
The invention can include an ellipse fitting subsystem for generating an ellipse to represent the upper torso of the occupant. That ellipse can then be used by the system to generate occupant characteristics for the tracking and predicting subsystem. In situations where a crash is detected, an at-risk-zone subsystem can be used to determine if the occupant is too close to the airbag for the airbag to safely deploy. If an occupant is not within the at-risk-zone, an impact assessment subsystem can be used to determine the impact that the airbag needs to absorb for a particular deployment so that the airbag can be set to deploy at the appropriate strength. In a preferred embodiment, the deployment strength of the airbag is set to absorb the kinetic energy of the occupant impacting into the airbag.
Various aspects of this invention will become apparent to those skilled in the art from the following detailed description of the preferred embodiment, when read in light of the accompanying drawings.
REFERENCES:
patent: 4179696 (1979-12-01), Quesinberry et al.
patent: 4625329 (1986-11-01), Ishikawa et al.
patent: 4985835 (1991-01-01), Sterler et al.
patent: 5051751 (1991-09-01), Gray
patent: 5074583 (1991-12-01), Fujita et al.
patent: 5229943 (1993-07-01), Eigler et al.
patent: 5256904 (1993-10-01), Tohbaru
patent: 5366241 (1994-11-01), Kithil
patent: 5398185 (1995-03-01), Omura
patent: 5413378 (1995-05-01), Steffens, Jr. et al.
patent: 5446661 (1995-08-01), Gioutsos et al.
patent: 5490069 (1996-02-01), Gioutsos et al.
patent: 5673365 (1997-09-01), Basehore et al.
patent: 5890085 (1999-03-01), Corrado et al.
patent: 5967549 (1999-10-01), Allen et al.
patent: 5983147 (1999-11-01), Krumm
patent: 6005958 (1999-12-01), Farmer et al.
patent: 6018693 (2000-01-01), Blackburn et al.
patent: 6026340 (2000-02-01), Corrado et al.
patent: 6198998 (2001-03-01), Farmer et al.
patent: 6272411 (2001-08-01), Corrado et al.
patent: 2 236 419 (1991-04-01), None
patent: 61-66905 (1994-06-01), None
patent: 61-66906 (1994-06-01), None
patent: WO02/30717 (2002-04-01), None
Greg Welch and Gary Bishop, “An Introduction to the Kalman Filter”, Sep. 4, 1997.
Cuchlinski Jr. William A.
Eaton Corporation
Marc McDieunel
Rader & Fishman & Grauer, PLLC
LandOfFree
Image processing system for detecting when an airbag should... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing system for detecting when an airbag should..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing system for detecting when an airbag should... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3170922