Image analysis – Applications – Range or distance measuring
Reexamination Certificate
2000-05-10
2004-05-04
Boudreau, Leo (Department: 2621)
Image analysis
Applications
Range or distance measuring
C382S104000
Reexamination Certificate
active
06731777
ABSTRACT:
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an optical object recognition system which detects objects in front of a vehicle such as an automobile, etc., using image capturing apparatus having cameras mounted on this vehicle. More specifically, the present invention concerns an object recognition system, which recognizes the characteristic features of objects using a plurality of windows in the captured images.
BACKGROUND OF THE INVENTION
In recent years, devices which determine the distance and size of objects in front of a vehicle, and which appropriately control the vehicle in accordance with this judgment, have been proposed for the purpose of improving the safety of vehicle operation.
Japanese Patent Application Kokai No. Hei 9-79821 describes one example of a device in which an optical distance measuring device consisting of two light-receiving elements is used to determine whether an object whose distance has been detected is a physical object or a road area (including characters or white lines on the road surface). The device calculates distance for respective calculation areas, and recognizes areas in which obstructions are present by clustering calculation areas whose mutual distances are within a fixed range and are proximate to each other in the horizontal direction. In forming the blocks, calculation areas whose distances have not been measured are also formed into blocks.
In the device disclosed in Japanese Patent Application Kokai No. Hei 9-79821, block formation is performed in cases where the distance values calculated from adjacent calculation areas are close to each other. Accordingly, in cases where a plurality of obstructions are located adjacent to each other in the image area, the obstructions may be formed into a single block and recognized as a single physical object. Because the ranges that are formed into blocks are fixed, a single obstruction may be erroneously recognized as a plurality of obstructions.
Accordingly, an object of the present invention is to provide a system which allows the accurate recognition of a plurality of physical objects present in the image area by utilizing information concerning past physical objects.
SUMMARY OF THE INVENTION
In accordance with one aspect of the invention, an object recognition system having one or more image sensors and a controller that is adapted for measuring the distance from the system to objects with respect to respective windows of an image captured by the sensors is provided. The controller forms clusters by uniting adjacent windows that have similar measured distances. The system includes a memory for storing data on a previously recognized physical object. The controller infers a physical object based on data on a previously recognized physical object and the speed of the vehicle relative to a previously recognized physical object and wherein said controller determines a combination of clusters that best matches the inferred physical object.
The controller groups the clusters into one or more cluster groups according to the distance from the vehicle carrying the system and selects from one of the cluster groups those clusters that overlaps with a physical object inferred by the controller. The controller matches combinations of the selected clusters with the inferred physical object based on the attributes of the combined clusters and the attributes of the inferred physical object.
According to another aspect of the invention, a method for recognizing a physical object is provided. The method comprises the steps of measuring distance from a vehicle to a physical object with respect to respective windows of an image captured by at least one image sensor, uniting adjacent windows that have similar measured distances to form clusters, inferring present position of a physical object based on data on a previously recognized physical object and the speed of the vehicle relative to the previously recognized physical, and determining a combination of clusters that best matches the inferred physical object.
The step of determining includes the step of selecting those clusters that overlaps with a physical object inferred by the controller and the step of matching combinations of the selected clusters with the inferred physical object based on the attributes of the combined clusters and the attributes of the inferred physical object.
According to another aspect of the invention, a method of recognizing a physical object is provided. The method comprises the steps of measuring distance from a vehicle to a physical object with respect to each window of an image captured one or more image sensors, clustering adjacent windows that are in a predetermined distance range, inferring present position of an physical object that was recognized in the previous recognition cycle, based on previous position of the physical object and the speed of the vehicle relative to the object, selecting clusters whose distance from the vehicle is within a predetermined tolerance relative to the distance of the inferred physical object and which overlaps with the inferred physical object, recognizing, as representing the physical object, the combined clusters that comprise one or more clusters selected by the selection step and that have closest attributes to the attributes of the inferred physical object.
The method may further comprise the steps of storing attributes of at least one sample physical object, and comparing attributes of combined clusters that were not selected in said step of selecting with attributes of said at least one sample physical object to recognize the sample physical object having closest attributes to be the physical object corresponding to the combined clusters.
The controller may comprise a micro-controller which typically includes a central processing unit (CPU) or a micro-processor, a read-only memory (ROM) containing control programs that when executed by the processor performs respective functions which are to be described hereafter. The controller also includes a random-access memory (RAM) that provides an working area for the CPU and temporary storage for various data and programs.
REFERENCES:
patent: 4908704 (1990-03-01), Fujioka et al.
patent: 5410346 (1995-04-01), Saneyoshi et al.
patent: 5554983 (1996-09-01), Kitamura et al.
patent: 5806019 (1998-09-01), Ishiyama
patent: 6035053 (2000-03-01), Yoshioka et al.
patent: 06-150196 (1994-05-01), None
patent: 09-079821 (1997-03-01), None
Nishigaki Morimichi
Saka Masakazu
Boudreau Leo
Honda Giken Kogyo Kabushiki Kaisha
Lahive & Cockfield LLP
Lu Tom Y.
LandOfFree
Object recognition 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 Object recognition system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Object recognition system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3260738