Image signal compression method and system
View Patent ↗An image signal compression method and system. Without involving the current processing pixel, an encryption key and a decryption key are respectively and separately generated by a compression subsystem and a decompression subsystem. Both of the encryption and decryption keys are separately generated using the same principle of pixel prediction. Both of the encryption and decryption keys have the same value. In the compression subsystem, the encryption key is subtracted from the pixel value of the current processing pixel to generate a compressed data. In the decompression subsystem, the decryption key is added to the compressed data to recover the pixel value of the current processing pixel.
1. A method for compressing and decompressing image signal, said method comprising:
a) a compression method for compressing a currently processing pixel X, said compression method comprising the steps of:
a1) determining first N pixel value differences respectively for N pairs of pixels belonging to first K neighboring pixels of said pixel X that have respectively first K currently buffered pixel values, wherein each of said first N pixel value differences is associated with one of said first K neighboring pixels;
a2) predicting a pixel Q to have similar pixel value with said pixel X by choosing said pixel Q from said K neighboring pixels, said pixel Q associated with one of said first pixel value differences that has the greatest value among said first N absolute pixel value differences; and
a3) compressing P(X) into a compressed data delta(X) to be transmitted for decompression, said delta(X)=P(X)−enKey(X), wherein said enKey(X) denotes an encryption key for compressing P(X) into said delta(X), said enKey(X) having value of P(Q), the pixel value of pixel Q.
2. The method of claim 1 , further comprising:
b) a decompression method for decompressing said delta(X) to recover the pixel value P(x) of a pixel x whose position within an image being compressed and decompressed is the same as the pixel position of said currently processing pixel X, said decompression method comprising the steps of:
b1) determining second N pixel value differences respectively for N pairs of pixels belonging to second K neighboring pixels of said pixel x that have respectively second K buffered pixel values, wherein each of said second N pixel value differences is associated with one of said second K neighboring pixels, said second N pixel value differences having numerical ranking equaling to that of said first N pixel value differences the pixel positions of said second K pixels equaling respectively to the pixel positions of said first K pixels, the pixel values of said second K neighboring pixels equaling respectively to the pixel values of said first K neighboring pixels;
b2) predicting a pixel q to have similar pixel value with said pixel x by choosing said pixel q from said second K neighboring pixels, said pixel q with its pixel value P(q) is associated with one of said second N pixel value differences that has the greatest value among said second N pixel value differences, said P(q) equaling to said P(Q); and
b3) decompressing said delta(X) into said P(x), said P(x)=delta(X)+deKey(x), wherein said deKey(x) denotes a decryption key for decompressing said delta(X) into P(x), said deKey(x) having value of P(q).
3. The method of claim 1 , wherein said pixel X has pixel position of (m,n)=(m-th row, n-th column) within said image, wherein said first K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein said first N pixel value differences are D 1 (m,n)=F(P(m−1,n−1), P(m−1−n)), D 2 (m,n)=F(P(m−1,n−1), P(m,n−1)), D 3 (m,n)=F(P(m,n−1), P(m−1,n)) and D 4 (m,n)=F(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
4. The method of claim 3 , wherein in response to D 1 (m,n) being greater than D 2 (m, n), D 3 (m, n) and D 4 (m, n), choose pixel (m−1,n) as said pixel Q, thereby setting said enKey(m, n)=P(m−1,n).
5. The method of claim 3 , wherein in response to D 2 (m,n) being greater than D 1 (m,n), D 3 (m,n) and D 4 (m,n), choose pixel (m,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m,n−1).
6. The method of claim 3 , wherein in response to D 3 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 4 (m,n), choose pixel (m−1,n+1) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n+1).
7. The method of claim 3 , wherein in response to D 4 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 3 (m,n), choose pixel (m−1,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n−1).
8. The method of claim 3 , wherein D 1 (m,n)=|P(m−1,n−1)−P(m−1−n)|, D 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, D 3 (m,n)=|P(m,n−1)−P(m−1,n)| and D 4 (m,n)=P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
9. The method of claim 3 , wherein D 1 (m,n)=P(m−1,n−1) 2 −P(m− 1 −n) 2 , D 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , D 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and D 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
10. The method of claim 2 , wherein said pixel x has pixel position of (m,n) within said image, wherein said second K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein said second N pixel value differences are d 1 (m,n)=f(P(m−1,n−1), P(m−1−n)), d 2 (m,n)=f(P(m−1,n−1), P(m,n−1)), d 3 (m,n)=f(P(m,n−1), P(m−1,n)) and d 4 (m,n)=f(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
11. The method of claim 10 , wherein in response to d 1 (m,n) being greater than d 2 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m−1,n) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n).
12. The method of claim 10 , wherein in response to d 2 (m,n) being greater than d 1 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m,n−1).
13. The method of claim 10 , wherein in response to d 3 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 4 (m,n), choose pixel (m−1,n+1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n+1).
14. The method of claim 10 , wherein in response to d 4 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 3 (m,n), choose pixel (m−1,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n−1).
15. The method of claim 10 , wherein d 1 (m,n)=|P(m−1,n−1)−P(m− 1 −n)|, d 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, d 3 (m,n)=|P(m,n−1)−P(m−1,n)| and d 4 (m,n)=|P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
16. The method of claim 10 , wherein d 1 (m,n)=P(m−1,n−1) 2 −P(m− 1 −n) 2 , d 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , d 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and d 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
17. A system for compressing and decompressing image signal said system comprises:
a compression subsystem comprising:
a first buffer adapted to store pixel values of first K neighboring pixels of a current processing pixel X;
a compression engine adapted for compressing said current processing pixel X by decomposing its pixel value P(X) into delta(X)+enKey(X), wherein delta(X) is the difference between P(X) and enKey(X), and wherein enKey(X) is the encryption key of said X, said encryption key enKey(X) is equal to P(Q), the pixel value of a pixel Q that is chosen from among said first K buffered neighboring pixels of said current processing pixel X; and
a decompression subsystem comprising:
a second buffer adapted to store pixel values of neighboring pixels of a pixel x whose pixel value is to be determined by decompressing said compressed data delta(X) transmitted by an image capturing device that compresses the pixel value P(X) of said pixel X having the same pixel position as said pixel x within an image being compressed and decompressed; and
a decompression module adapted to assign to said x a pixel value P(x) by decompressing delta(X) into delta(X)+deKey(x), said decryption key deKey(x) is P(q), the pixel value of a pixel q that is chosen from among said second K buffered neighboring pixel values of pixel x, wherein said pixel X has pixel position of (m,n)=(m-th row, n-th column) within said image, wherein said first K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein first N pixel value differences are D 1 (m,n)=F(P(m−1,n−1), P(m−1−n)), D 2 (m,n)=F(P(m−1,n−1), P(m,n−1)), D 3 (m,n)=F(P(m,n−1), P(m−1,n)) and D 4 (m,n)=F(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
18. The system of claim 17 , wherein in response to D 1 (m,n) being greater than D 2 (m,n), D 3 (m,n) and D 4 (m,n), choose pixel (m−1,n) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n).
19. The system of claim 17 , wherein in response to D 2 (m,n) being greater than D 1 (m,n), D 3 (m,n) and D 4 (m,n), choose pixel (m,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m,n−1).
20. The system of claim 17 , wherein in response to D 3 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 4 (m,n), choose pixel (m−1,n+1) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n+1).
21. The system of claim 17 , wherein in response to D 4 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 3 (m,n), choose pixel (m−1,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n−1).
22. The system of claim 17 , wherein D 1 (m,n)=|P(m−1,n−1)−P(m−1−n)|, D 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, D 3 (m,n)=|P(m,n−1)−P(m−1,n)| and D 4 (m,n)=|P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
23. The system of claim 17 , wherein D 1 (m,n)=P(m−1,n−1) 2 −P(m−1−n) 2 , D 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , D 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and D 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
24. The system of claim 17 , wherein said pixel x has pixel position of (m,n) within said image, wherein said second K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein said second N pixel value differences are d 1 (m,n)=f(P(m−1,n−1), P(m−1−n)), d 2 (m,n)=f(P(m−1,n−1), P(m,n−1)), d 3 (m,n)=f(P(m,n−1), P(m −1 ,n)) and d 4 (m,n)=f(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n), wherein said d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n) have the same numerical ranking as that of D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
25. The system of claim 24 , wherein in response to d 1 (m,n) being greater than d 2 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m−1,n) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n).
26. The system of claim 24 , wherein in response to d 2 (m,n) being greater than d 1 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m,n−1).
27. The system of claim 24 , wherein in response to d 3 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 4 (m,n), choose pixel (m−1,n+1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n+1).
28. The system of claim 24 , wherein in response to d 4 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 3 (m,n), choose pixel (m−1,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n−1).
29. The system of claim 24 , wherein d 1 (m,n)=P(m−1,n−1)−P(m−1−n)|, d 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, d 3 (m,n)=|P(m,n−1)|P(m−1,n)| and d 4 (m,n)=|P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
30. The system of claim 24 , wherein d 1 (m,n)=P(m−1,n−1) 2 −P(m−1−n) 2 , d 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , d 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and d 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
31. An image capturing device comprising:
a buffer adapted to store pixel values of K neighboring pixels of a current processing pixel X; and
a compression engine adapted for compressing said current processing pixel X by decomposing its pixel value P(X) into delta(X)+enKey(X), wherein delta(X) is the difference between P(X) and wherein enKey(X) is the encryption key of said X, said encryption key enKey(X) is equal to P(Q), the pixel value of a pixel Q that is chosen from among said K buffered neighboring pixels of said current processing pixel X, wherein said pixel Q is chosen based on a prediction that among said buffered neighboring pixel values of pixel X, P(Q) is similar in value to P(X), wherein said prediction does not require P(X).
32. The image capturing device of claim 31 , wherein said pixel X has pixel position of (m,n)=(m-th row, n-th column) within said image, wherein said first K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein first N pixel value differences are D 1 (m,n)=F(P(m−1,n−1), P(m−1−n)), D 2 (m,n)=F(P(m−1,n−1), P(m,n−1)), D 3 (m,n)=F(P(m,n−1), P(m−1,n)) and D 4 (m,n)=F(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
33. The image capturing device of claim 32 , wherein in response to D 1 (m,n) being greater than D 2 (m,n), D 3 (m,n) and D 4 (m,n), choose pixel (m−1,n) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n).
34. The image capturing device of claim 32 , wherein in response to D 2 (m,n) being greater than D 1 (m,n), D 3 (m,n) and D 4 (m,n), choose pixel (m,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m,n−1).
35. The image capturing device of claim 32 , wherein in response to D 1 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 4 (m,n), choose pixel (m−1,n+1) as said pixel Q, thereby setting said enKey(m,n)=P(m − 1,n+1).
36. The image capturing device of claim 32 , wherein in response to D 1 (m,n) being greater than D 1 (m,n), D 2 (m,n) and D 3 (m,n), choose pixel (m−1,n−1) as said pixel Q, thereby setting said enKey(m,n)=P(m−1,n−1).
37. The image capturing device of claim 32 , wherein D 1 (m,n)=|P(m−1,n−1)−P(m−1−n)|, D 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, D 3 (m,n)=|P(m,n−1)|P(m−1,n)| and D 4 (m,n)=|P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
38. The image capturing device of claim 32 , wherein D 1 (m,n)=P(m−1,n−1) 2 −P(m−1−n) 2 , D 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , D 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and D 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with D 1 (m,n), D 2 (m,n), D 3 (m,n) and D 4 (m,n).
39. An apparatus for performing decompression of image signal, said apparatus comprises:
a buffer adapted to store pixel values of K neighboring pixels of a pixel x whose pixel value is to be determined by decompressing a compressed data transmitted by an image capturing device that compresses into a compressed data delta(X) the pixel value P(X) of a pixel X having the same pixel position as said pixel x within an image being compressed and decompressed; and
a decompression module adapted to assign to said x a pixel value P(x) by decompressing said delta(X) into delta(X)+deKey(x), said decryption key deKey(x) is P(q), the pixel value of a pixel q that is chosen from among said K buffered neighboring pixel values of pixel x, wherein said pixel x has pixel position of (m,n) within said image, wherein said second K neighboring pixels have pixel positions (m−1, n−2), (m−1,n−1), (m−1, n), (m−1, n+1), and (m, n−1), wherein said second N pixel value differences are d 1 (m,n)=f(P(m−1,n−1), P(m−1−n)), d 2 (m,n)=f(P(m−1,n−1), P(m,n−1)), d 3 (m,n)=f(P(m,n−1), P(m−1,n)) and d 4 (m,n)=f(P(m,n−1), P(m−1,n−2)), wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
40. The apparatus of claim 39 , wherein in response to d 1 (m,n) being greater than d 2 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m−1,n) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n).
41. The apparatus of claim 39 , wherein in response to d 2 (m,n) being greater than d 1 (m,n), d 3 (m,n) and d 4 (m,n), choose pixel (m,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m,n−1).
42. The apparatus of claim 39 , wherein in response to d 3 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 4 (m,n), choose pixel (m−1,n+1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n+1).
43. The apparatus of claim 39 , wherein in response to d 4 (m,n) being greater than d 1 (m,n), d 2 (m,n) and d 3 (m,n), choose pixel (m−1,n−1) as said pixel q, thereby setting said deKey(m,n)=P(m−1,n−1).
44. The apparatus of claim 39 , wherein d 1 (m,n)=|P(m−1,n−1)−P(m−1−n)|, d 2 (m,n)=|P(m−1,n−1)−P(m,n−1)|, d 3 (m,n)=|P(m,n−1)−P(m−1,n)| and d 4 (m,n)=|P(m,n−1)−P(m−1,n−2)|, wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).
45. The apparatus of claim 39 , wherein d 1 (m,n)=P(m−1,n−1) 2 −P(m−1−n) 2 , d 2 (m,n)=P(m−1,n−1) 2 −P(m,n−1) 2 , d 3 (m,n)=P(m,n−1) 2 −P(m−1,n) 2 and d 4 (m,n)=P(m,n−1) 2 −P(m−1,n−2) 2 , wherein pixels (m−1,n), (m,n−1), (m−1,n+1) and (m−1,n−1) are associated respectively with d 1 (m,n), d 2 (m,n), d 3 (m,n) and d 4 (m,n).