Data processing: vehicles – navigation – and relative location – Relative location – Collision avoidance
Reexamination Certificate
2000-06-09
2001-12-04
Beaulieu, Yonel (Department: 3661)
Data processing: vehicles, navigation, and relative location
Relative location
Collision avoidance
C701S302000, C701S028000, C701S070000, C701S223000, C348S025000, C348S113000, C348S118000, C340S435000, C340S436000, C340S901000
Reexamination Certificate
active
06327536
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a vehicle environment monitoring system that monitors an environment of an automotive vehicle on which the vehicle is installed. The system detects an object which may collide with the vehicle from an image obtained by an imaging device mounted on the vehicle, thereby avoiding collision against large animals, such as a deer, a bear and the like, since such collision has an adverse influence on the running of the vehicle.
2. Description of the Prior Art
Conventionally, a detection system for detecting an object crossing a road ahead of a vehicle was proposed in Japanese Laid-Open Patent Publication (Kokai) No. 9-226490, the proposed system detects a distance between the vehicle and the object based on the displacement between images of the object, that is, the parallax of images obtained by two CCD (charge-coupled device) cameras mounted on the vehicle, thereby detecting a pedestrian 30 to 60 meters ahead of the vehicle.
However, the above conventional apparatus determines the possibility of collision against an object by using the direction of an optical flow of the object detected based on images of the object obtained by the cameras, without further processing, which results in degraded determining accuracy depending on a relative distance between the vehicle and the object, and the speed of the vehicle. That is, when a vehicle speed is much higher than the speed of a moving object, an optical flow on the image, which corresponds e.g. to an object actually moving toward the center of a road, is sometimes tracked as a vector extending outside the road, which causes an erroneous determination of the possibility of potential collision.
SUMMARY OF THE INVENTION
It is an object of the invention to provide a vehicle environment monitoring system, which is capable of more accurately detecting the movement of an object existing in an environment of the vehicle, and determining the possibility of collision between the object and the vehicle, for thereby more appropriately warning the driver of such possibility.
To attain the above object, the present invention provides a vehicle environment monitoring system for an automotive vehicle, which detects an object existing in an environment of the automotive vehicle from an image obtained by imaging means mounted on the automotive vehicle.
The vehicle environment monitoring system is characterized by comprising:
relative position-detecting means for detecting a relative position of the object to the automotive vehicle from the image obtained by the imaging means to obtain position data;
movement vector-calculating means for calculating positions of the object in a real space based on a plurality of time series items of the position data detected on the object by the relative position-detecting means, and calculating a movement vector of the object based on the positions in the real space; and
determining means for determining whether or not the object has a high possibility of collision against the automotive vehicle based on the movement vector.
According to this vehicle environment monitoring system, the relative position of the object to the automotive vehicle is detected from the image obtained by the imaging means to obtain position data, the and the positions of the object in a real space are calculated based on a plurality of time series items of the position data detected on the object. Then, a movement vector of the object is calculated based on the positions of the object in the real space, and it is determined whether or not the object has a high possibility of collision against the automotive vehicle based on the movement vector. Therefore, differently from the conventional system, it is possible to prevent the occurrence of an erroneous determination, and thereby enhance accuracy of the determination of the possibility of collision.
Preferably, the movement vector-calculating means includes approximate straight line-calculating means for calculating an approximate straight line approximating a locus of relative movement of the object, and position data-correcting means for correcting the time series items of the position data by using the approximate straight line, and the movement vector-calculating means calculates the movement vector based on the corrected time series items of the position data.
According to this preferred embodiment, an approximate straight line approximating a locus of relative movement of the object is calculated, and the time series items of the position data are corrected by using the approximate straight line. The movement vector is calculated based on the corrected time series items of the position date. Therefore, it is possible to reduce adverse influence of errors in detection of the position data, and more accurately estimate the possibility of collision against the object in advance.
Preferably, the determining means carries out the determination by applying conditions for determining a potential collision, the conditions being dependent on a width of the vehicle.
According to this preferred embodiment, the possibility of collision is determined by applying conditions for determining a potential collision, which are dependent on the width of the vehicle. This makes it possible to more accurately determine the possibility of collision and prevent the driver from being unnecessarily warned.
More preferably, the approximate straight line-calculating means calculates the approximate straight line, as a three-dimensional straight line including data of a height of the object.
According to this preferred embodiment, it is possible to accurately determine the possibility of collision, even when the driver is driving on a hilly road.
Preferably, the imaging means comprises two infrared cameras capable of detecting infrared rays.
According to this preferred embodiment, an animal, running automotive vehicle, and the like can easily be detected, even when the driver is driving at night and it is difficult to recognize them.
Preferably, the imaging means comprises two TV cameras for detecting infrared rays or visible rays, and the relative position-detecting means includes search area-setting means for setting, based on a position of an object image contained in an image output from one of the two TV cameras, a search area within an image output from the other of the two TV cameras for searching for a corresponding object image contained in the image output from the other of the two TV cameras, corresponding object image-identifying means for identifying the corresponding object image by carrying out a correlation operation on data within the search area, and distance-calculating means for calculating a distance from the vehicle to the object based on a parallax between the object image and the corresponding object image.
More preferably, the data within the search area is gray scale data containing halftone information.
According to this preferred embodiment, by using not binary data but gray scale data, it is possible to effect more accurate correlation operation, thereby preventing erroneous identification of a corresponding object image.
Further preferably, the relative position-detecting means includes tracking means for tracking an object moving in the image obtained by the imaging means, by using binary data formed by binarizing the gray scale data.
Even more preferably, the tracking means recognizes the object being tracked by using run length data encoded from the binary data.
According to this preferred embodiment, it is possible to reduce the amount of data concerning the object being tracked, thereby saving the capacity of the memory device.
Still more preferably, the tracking means determines an identity of the object being tracked, based on position coordinates of a centroid of an image thereof, an area of the image thereof, and an aspect ratio of a rectangle circumscribing the image thereof.
According to this preferred embodiment, it is possible to accurately determine the identity of an object being tracked.
Pre
Hattori Hiroshi
Nagaoka Nobuharu
Tsuji Takayuki
Watanabe Masahito
Beaulieu Yonel
Blackman William D.
Carrier Joseph P.
Carrier Blackman & Associates
Honda Giken Kogyo Kabushiki Kaisha
LandOfFree
Vehicle environment monitoring system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Vehicle environment monitoring system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Vehicle environment monitoring system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2593789