Optical image encoding and comparing using scan autocorrelation

• US 4,745,633 A
• Filed: 08/18/1986
• Issued: 05/17/1988
• Est. Priority Date: 08/18/1986
• Status: Expired due to Fees

• Current Assignee: Ernest Greene, Peter Waksman
• Original Assignee: Ernest Greene, Peter Waksman
• Inventors: Waksman, Peter, Greene, Ernest
• Primary Examiner(s): Groody, James J.
• Assistant Examiner(s): Parker, Michael D.
• Application Number:
  US06/897,780
• Time in Patent Office: 638 Days
• Field of Search: 382/25, 382/36, 382/42, 382/56, 382/, 364/604, 364/728, 364/819, 358/260, 358/261
• US Class Current: 382/243
• CPC Class Codes: G06V 10/421 by analysing segments intersecting the pattern

Abstract

A method for encoding shape and pattern information from the image (20) of an electronic camera (10). The image is converted into a discrete array of values (30), and the values in the cross sections of the array are sampled (circuit 40), marked for transitions (circuit 50), and the transition markers are then autocorrelated by circuit 60 to form an autocorrelation function for each cross section. The autocorrelation functions corresponding to each cross section are then combined by circuit 70 into a global summary function, which provides a quantitative measure of the information contained in the image. The global summary functions which are derived from images may be used to judge their similarity or otherwise compare and classify them.

Claims

1. A system for encoding information from the optical image of a camera, comprising,

(a) means for representing a predetermined characteristic of said image by an array of values,

(b) means for sampling an ordered sequence of the values of said array, and for marking transitions in said ordered sequence by a sequence of marker values, and

(c) means for autocorrelating said sequence of marker values to provide an autocorrelation function for said sequence of marker values, and thus for said ordered sequence.

2. The encoding system of claim 1, further including means for storing said values of said array, said sequence of marker values, and the values of said autocorrelation function.

3. The encoding system of claim 1 wherein said means for sampling, said means for marking transitions, and said means for autocorrelating said sequence of marker values comprises a hard-wired, direct-logic network.

4. The encoding system of claim 1 wherein said means for sampling said ordered sequence of values of said array is arranged to provide said ordered sequence as a temporal signal, and wherein said means for marking transitions and for autocorrelation of said sequence of marker values comprises a plurality of shift registers and gates which process said temporal signal.

5. The encoding system of claim 1 wherein said means for autocorrelating said sequence of marker values comprises a computer and an instructional program for said computer.

6. The encoding system of claim 1, further including means for preselecting a plurality of said ordered sequences and for calculating a corresponding plurality of autocorrelation functions, wherein said means for marking transitions is applied to each member of said plurality of ordered sequences and said means for autocorrelating is applied to each sequence of marker values thus derived, whereby an autocorrelation function is generated for each member of said plurality of ordered sequences.

7. The encoding system of claim 6 wherein said means for representing predetermined characteristics of said image is arranged to generate an array which has an hexagonal shape, and said means for preselecting said plurality of ordered sequences is arranged to preselect those sequences which are the natural cross sections of said hexagonal array in three directions.

8. The encoding system of claim 6 wherein said means for preselecting said plurality of ordered sequences is arranged to select said sequences from a sector of said array, whereby the information contained at specific locations in said image may be isolated and selectively encoded.

9. The encoding system of claim 6, further including means for adding a set of specific location markers to said sequence of marker values, whereby the information contained at specific locations in said image may determine the values of said autocorrelation functions.

10. The encoding system of claim 6, further including means for algebraically combining said plurality of autocorrelation functions, so as to provide a single summary function for said array and thus for said image.

11. The encoding system of claim 10, further including means for storing said summary function.

12. The encoding system of claim 10 wherein said means for algebraically combining said plurality of autocorrelation functions is arranged to calculate the sum of said plurality of autocorrelation functions.

13. The encoding system of claim 10, further including means for repeating said encoding and combining process, thereby to accommodate applications in which said image array moves in relation to said camera.

14. The encoding system of claim 10, further including means for sampling a plurality of said images, and for the repeated application of said means for preselecting a plurality of ordered sequences for each said image, of said means for calculating said corresponding plurality of autocorrelation functions for each said image, of said means of combining said plurality of autocorrelation functions into a single summary function for each said image, and for storing said summary functions.

15. The encoding system of claim 14, further including means for judging similarity among said images by the calculation of a comparison value for each pair of said summary functions, whereby the relative magnitude of each comparison value defines similarity among said images.

16. The encoding system of claim 15 wherein said means for judging is accomplished by the calculation of a comparison value which is the mean square difference between the summary functions.

17. The encoding system of claim 15 wherein said means for judging is accomplished by the calculation of a comparison value which is the absolute difference between the cumulative forms of said summary functions.

18. The encoding system of claim 6, further including means for calculating and storing a set of moments for said plurality of autocorrelation functions, means for averaging said set of moments to produce a set of summary moments, and means for classifying the information in said array using the relative magnitude of elements in said set of summary moments.

19. A system for encoding and combining information from the optical image of a camera, comprising,

(a) means for representing predetermined characteristics of said image by an array of values,

(b) means for preselecting a plurality of ordered sequences of values of said array,

(c) means for sampling each member of said plurality of ordered sequences, and for autocorrelating each member of said plurality to produce a coresponding plurality of autocorrelation functions, and

(d) means for algebraically combining said plurality of autocorrelation functions into a single summary function for said array and thus for said image.

20. The enooding system of claim 19, further including means for sampling a plurality of said images, means for preselecting said plurality of ordered sequences for each said image, means for calculating said plurality of autocorrelation functions for each said image, means for calculating said single summary function for each said image, means for storing said summary functions, and means for comparing and establishing the degree of quantative similarity among said summary functions by the calculation of a comparison value.