Machine vision camera and video reprocessing system

• US 5,438,360 A
• Filed: 07/28/1994
• Issued: 08/01/1995
• Est. Priority Date: 09/08/1992
• Status: Expired due to Fees

• Current Assignee: Active Motion Systems LLC
• Original Assignee: Paul Howard Mayeux
• Inventors: Edwards, Michael K.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): Shalwala, Bipin
• Application Number:
  US08/282,169
• Time in Patent Office: 369 Days
• Field of Search: 348/208, 348/143, 348/154, 348/155, 348/416
• US Class Current: 348/208.4
• CPC Class Codes: H04N 23/68 for stable pick-up of the scene, e.g. compensating for camera body vibrations
• CPC Class Codes: H04N 23/6811 based on the image signal
• CPC Class Codes: H04N 23/683 performed by a processor, e.g. controlling the readout of an image memory
• CPC Class Codes: H04N 25/00 Circuitry of solid-state image sensors [SSIS]; Control thereof

Abstract

An improved camera system intended for a machine vision system which produces an image output that is specially adapted to be received by a machine vision system. In one embodiment, the camera system receives components from three portions of the electromagnetic spectrum, these being infrared, visible, and ultraviolet, and then combines these three portions essentially as expanded colors in an expanded version of an RGB "red/green/blue" video system. This provides a better image in all types of weather and visibility conditions. In addition, the camera system includes dynamic range transform logic which selectively compresses the dynamic range of the respective pixel values comprising the image to extract the most useful information from the extended dynamic range of the image. This reduces the amount of information required to be processed while retaining a quality image. The camera system also includes motion compensation logic that removes the motion or jitter caused by the camera system being subject to vibration from various environmental influences. The system also optionally removes stationary background features from the output image.

Claims

1. A video preprocessing system for removing effects of camera motion, comprising:

a video source for providing incoming signals representing an image; and

a motion compensator coupled to said video source receiving said incoming image for adjusting said incoming image to compensate for camera motion caused by environmental influence on the camera,

said motion compensator comprising;

a background frame buffer for storing an accumulated background image of said image; and

a comparator coupled to said background frame buffer which compares values of a plurality of incoming pixels with values of selected background pixels, wherein each of said background pixels that are compared with said incoming pixels are displaced in position from said incoming pixels by one of a plurality of predetermined relative displacements, wherein said incoming image comprises a plurality of blocks each comprising a first number of pixels, and a different predetermined relative displacement from a first number of predetermined relative displacements is used for each pixel comprising one of said blocks wherein said comparison of said incoming pixels with said background pixels generates error values;

an error matrix store for storing said error values; and

determining means coupled to said error matrix store for determining a total displacement that can be made to said incoming image to compensate for said camera motion, wherein said determining means uses said error values in said error matrix store in determining said total displacement;

wherein said incoming image is adjusted according to said total displacement to compensate for said camera motion.

2. The video preprocessing system of claim 1, wherein said first number of relative displacements are applied to said first number of pixels comprising each of said blocks according to a predetermined pattern, said pattern repeating for each of said blocks.

3. The video preprocessing system of claim 1, wherein said error matrix store comprises a first number of locations corresponding to said first number of relative displacements for accumulating said error values associated with each of said relative displacements.

4. The video preprocessing system of claim 3, wherein said determining means analyzes said error values in said error matrix store to determine a relative displacement that best compensates for said environmental influence on said camera system.

5. The video preprocessing system of claim 4 wherein said error matrix store comprises rows and columns of said first number of locations:

wherein said determining means sums said accumulated error values in each of said rows and columns comprising said error matrix store, determines the row and column with the minimum sum of accumulated error values, and determines the location storing the minimum accumulated error value within a region surrounding an intersection of said minimum sum row and column.

6. The video preprocessing system of claim 1,

wherein said predetermined relative displacements are made from an initial displacement guess;

and

wherein said total displacement comprises a selected relative displacement in addition to said initial displacement guess.

7. The video preprocessing system of claim 1, said motion compensator further comprising:

an incoming pixel store receiving said incoming pixels comprising said incoming image from said video source; and

a coordinate generator coupled to said background frame buffer for generating coordinates of displaced background pixels to be compared with said incoming pixels;

wherein said comparator receives said incoming pixels and compares said incoming pixel values with said displaced background pixel values to produce said error values.

8. The video preprocessing system according to claim 1, wherein said background frame buffer stores an accumulated background image comprising historically stationary objects.

9. The video preprocessing system according to claim 8, wherein said background frame buffer stores an accumulated background image according to the equation:

space="preserve" listing-type="equation">B.sub.new →

B·

(1-C)+I·

C

wherein B_new is the new accumulated background image, B is the previous accumulated background image, I is the incoming image, and C is a value less than one.

10. The video preprocessing system according to claim 9, wherein said background frame buffer stores an accumulated background image according to the equation:

space="preserve" listing-type="equation">B.sub.new →

B·

(1-C)+I·

C

wherein B_new is tile new accumulated background image, B is the previous accumulated background image, I is the incoming image, and C is all exponential decay factor.

11. A method for adjusting an image output from a video source to remove jitter caused by external influences on the video source, the method comprising the steps of:

(a) receiving an incoming pixel from a video source, said incoming pixel being part of an incoming image;

(b) computing coordinates of a displaced background pixel to be compared with said incoming pixel, said coordinates of said displaced background pixel being based on a relative displacement;

(c) fetching said displaced background pixel;

(d) computing an error value based on values of said incoming pixel and said displaced background pixel after said step of fetching;

(e) accumulating said error value with prior error values computed using said relative displacement;

(f) repeating steps (a) through (e) for a plurality of incoming pixels comprising said incoming image using a plurality of different relative displacements, wherein said incoming image comprises a plurality of blocks each comprising a first number of pixels and a different relative displacement chosen from a first number of relative displacements is used for each pixel comprising one of said blocks;

(g) examining said accumulated error values for each of said relative displacements to determine a relative displacement that compensates for said jitter after step (f); and

(h) adjusting said incoming image according to said determined relative displacement to compensate for said jitter.

12. The method of claim 11, wherein said step of computing coordinates computes coordinates of said displaced background pixel based on one of said plurality of relative displacements and an initial displacement guess;

wherein said step of adjusting adjusts said image according to said initial displacement guess and said determined relative displacement.

13. The method of claim 11, wherein said first number of relative displacements are applied to said first number of pixels comprising each of said blocks according to a pattern, said pattern repeating for each of said blocks.

14. The method of claim 11, wherein said step of accumulating comprises accumulating said error values associated with each of said relative displacements in an error matrix store.

15. The method of claim 11, wherein said step of examining includes summing said accumulated error values in respective rows and columns of said error matrix store and determining the row and column with the minimum sum of accumulated error values.