Apparatus and method for vehicle following with dynamic feature recognition

• US 5,835,880 A
• Filed: 07/19/1995
• Issued: 11/10/1998
• Est. Priority Date: 07/19/1995
• Status: Expired due to Term

• Current Assignee: VI&T Group Incorporated
• Original Assignee: VI&T Group Incorporated
• Inventors: Wang, Wei-Ping, Zhang, Ruiming, Tang, Oing, Gan, Zhongxue, Moh, John
• Primary Examiner(s): NGUYEN, TAN QUANG
• Application Number:
  US08/503,943
• Time in Patent Office: 1,210 Days
• Field of Search: 364/447, 364/460, 364/461, 364/462, 364/561, 340/903, 340/435, 340/937, 382/103, 382/104, 382/154, 382/199, 382/266, 382/283, 382/174, 356/3.14, 701/205, 701/300, 701/301, 701/302
• US Class Current: 701/205
• CPC Class Codes: G01S 5/16 using electromagnetic waves other than radio waves
• CPC Class Codes: G05D 1/0251 extracting 3D information from a plurality of images taken from different locations, e.g. stereo vision stereoscopic image analysis H04N13/00; depth recovery from images G06T7/593
• CPC Class Codes: G06T 7/593 from stereo images
• CPC Class Codes: G06V 10/245 by locating a pattern; Special marks for positioning
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
• CPC Class Codes: G06V 20/56 exterior to a vehicle by using sensors mounted on the vehicle
• CPC Class Codes: G06V 2201/08 Detecting or categorising vehicles

Abstract

In an apparatus and method for automatic vehicle following, a pair of video cameras are mounted on the front portion of a following vehicle, and each camera generates image data indicative of a respective image of a distinctive mark located on a rearward portion of a lead vehicle. An image processor generates two one-dimensional image-intensity distributions for each mark image based on the image data, and in turn generates a respective edge-intensity distribution based on each image-intensity distribution. The edge-intensity distributions are each generated by taking the first derivative of the respective image-intensity distribution. The image processor in turn generates a respective feature vector based on each edge-intensity distribution, wherein the vector components of each feature vector define select characteristics of the respective mark image. The center points of the two corresponding mark images are determined based on the feature vectors, and a signal processor generates signals indicative of the following distance and the heading angle of the lead vehicle based on the image center points. A window enclosing each mark image is moved upon generating a new mark image to the approximate center point of the preceding image, to substantially follow image movements from one camera image frame to the next, and thereby suppress noise and enhance the image-processing speed of the apparatus.

Claims

1. An apparatus for controlling a following vehicle to substantially follow the path of a lead moving object, comprising:

a distinctive mark on the lead moving object;

first means for generating image data indicative of two corresponding images of the mark on the lead moving object;

second means for generating at least one image profile for each mark image based on average values for a plurality of groups of image data points to further suppress noise in the image profiles;

third means for generating at least one characteristic feature of each mark image selected from the group including;

(i) the maximum slope of the respective image profile, (ii) the minimum slope of the respective image profile, and (iii) at least one dimension defining the distance between two edges of the respective image profile; and

fourth means for generating signals indicative of at least one of (i) the approximate distance between the lead moving object and the following vehicle and (ii) the heading angle of the lead moving object based on the characteristic features of the at least two image profiles.

2. An apparatus as defined in claim 1, wherein each image profile is comprised of at least one image-intensity profile generated by compressing the image data to eliminate noise and intensify the image signals.

3. An apparatus as defined in claim 2, wherein the image-intensity profiles are generated by compressing the image data for each mark image in first and second coordinate directions and in turn generating two one-dimensional intensity distributions for each mark image.

4. An apparatus as defined in claim 3, wherein each one-dimensional intensity distribution is generated by summing intensity values of the image in the respective coordinate direction.

5. An apparatus as defined in claim 1, wherein the third means for generating at least one characteristic feature of each mark image generates a respective edge-intensity distribution based on each image profile.

6. An apparatus as defined in claim 5, wherein the third means for generating at least one characteristic feature generates at least one feature vector based on characteristics of the respective edge-intensity distribution.

7. An apparatus as defined in claim 6, wherein the at least one feature vector has vector components based on the at least one of (i) the maximum slope of the respective image profile, (ii) the minimum slope of the respective image profile, and (iii) at least one dimension defining the distance between two edges of the respective image profile.

8. An apparatus as defined in claim 6, wherein the third means for generating at least one characteristic feature compares each feature vector to a corresponding predetermined feature vector to determine if the respective mark image substantially matches the leading object'"'"'s mark.

9. An apparatus as defined in claim 1, wherein the fourth means generates signals indicative of common points in the two corresponding mark images based on the characteristic features.

10. An apparatus as defined in claim 9, wherein the common points are each approximately the center point of the respective mark image.

11. An apparatus as defined in claim 9, wherein the fourth means generates signals indicative of (i) the distance between the leading object and the following vehicle and (ii) the heading angle of the leading object, based on the coordinate positions of the common points in the two corresponding mark images.

12. An apparatus as defined in claim 1, further comprising means for generating a moving window enclosing each mark image within a respective image frame.

13. An apparatus as defined in claim 12, wherein each window is moved upon generating a new mark image to the approximate center point of the preceding image to substantially follow image movements.

14. An apparatus as defined in claim 1, wherein the distinctive mark is defined by a geometric configuration formed by a reflective surface and substantially surrounded by a substantially non-reflective surface.

15. An apparatus as defined in claim 1, wherein the lead moving object is a vehicle.

16. A method for controlling a following vehicle to substantially follow the path of a lead moving object, comprising the steps of:

generating image data indicative of two corresponding images of a distinctive mark on the lead moving object;

generating at least one image profile for each mark image based on average values for a plurality of groups of image data points to further suppress noise in the image profiles;

generating at least one characteristic feature of each mark image selected from the group including (i) the maximum slope of the respective image profile, (ii) the minimum slope of the respective image profile, and (iii) at least one dimension defining the distance between two edges of the respective image profile; and

generating signals indicative of at least one of (i) the approximate distance between the lead moving object and the following vehicle and (ii) the heading angle of the lead moving object based on the characteristic features of the at least two image profiles.

17. A method as defined in claim 16, wherein the step of generating at least one image profile includes generating at least one image-intensity profile by compressing the image data to eliminate noise and intensify the image signals.

18. A method as defined in claim 17, wherein the step of generating at least one image profile further includes compressing the image data for each mark image in first and second coordinate directions and in turn generating two one-dimensional intensity distributions for each mark image.

19. A method as defined in claim 18, further comprising the step of generating each one-dimensional intensity distribution by summing intensity values of the image in the respective coordinate direction.

20. A method as defined in claim 16, further comprising the step of generating the at least one characteristic feature by generating a respective edge-intensity distribution based on each image profile.

21. A method as defined in claim 20, further comprising the step of generating at least one feature vector based on characteristics of each edge-intensity distribution.

22. A method as defined in claim 21, wherein the at least one feature vector has vector components based on the at least one of (i) the maximum slope of the respective image profile, (ii) the minimum slope of the respective image profile, and (iii) at least one dimension defining the distance between two edges of the respective image profile.

23. A method as defined in claim 21, further comprising the step of comparing each feature vector to a corresponding predetermined feature vector to determine if the respective mark image substantially matches the leading object'"'"'s mark.

24. A method as defined in claim 16, further comprising the steps of generating signals indicative of common points in the two corresponding mark images based on the characteristic features, and generating signals indicative of (i) the distance between the leading object and the following vehicle and (ii) the heading angle of the leading object, based on the coordinate positions of the common points in the two corresponding mark images.

25. A method as defined in claim 24, further comprising the step of generating a moving window enclosing each mark image within a respective image frame.

26. A method as defined in claim 16, wherein the lead moving object is a vehicle.