Fuzzy logic based color transition improvement method and system
First Claim
1. A computer implemented method stored on a recordable media for performing digital color transition improvement (DCTI) on a N-pixel delayed signal of a digital chrominance signal, said method comprising.using said digital chrominance signal, said N-pixel delayed signal and a 2N-pixel delayed signal of said digital chrominance signal, generating a first and second difference signals of first order;
- using said first and second difference signals of first order, generating a difference signal of second order; and
generating a first weighted sum of the digital chrominance signal, the 2N-pixel delayed signal and a first weight factor determined by applying a first set of fuzzy logic inference rules to the first difference signal of first order and the difference signal of second order.
4 Assignments
0 Petitions
Accused Products
Abstract
A fuzzy logic based digital color transition improvement method and apparatus for increasing color sharpness by replacing the slow transition color edges with edges that have steeper rising and falling times. Fuzzy logic used here to decide where the transition happens and how to enhance the signal when transition happens. Based on the results of fuzzy logic inference, through weighting among input digital color signal, its N-pixel delayed signal and 2N-pixel delayed signal, the output signal has steep and smooth color edges without ringing.
23 Citations
33 Claims
-
1. A computer implemented method stored on a recordable media for performing digital color transition improvement (DCTI) on a N-pixel delayed signal of a digital chrominance signal, said method comprising.
using said digital chrominance signal, said N-pixel delayed signal and a 2N-pixel delayed signal of said digital chrominance signal, generating a first and second difference signals of first order; -
using said first and second difference signals of first order, generating a difference signal of second order; and
generating a first weighted sum of the digital chrominance signal, the 2N-pixel delayed signal and a first weight factor determined by applying a first set of fuzzy logic inference rules to the first difference signal of first order and the difference signal of second order. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12)
generating a second weighted sum of the first weighted sum, the N-pixel delayed signal and a second weight factor determined by applying a second set of fuzzy logic inference rules to the second difference signal of first order and the first difference signal of first order, wherein said second weighted sum has an enhanced digital color transition in comparison to the digital chrominance signal.
-
-
3. The method of claim 2, wherein the second weighted sum is characterized as:
(the second factor) * (the N-pixel delayed signal)+(1−
the second weight factor) * (the first weighted sum).
-
4. The method of claim 2, comprising:
generating the second weight factor by performing a second fuzzy logic inference algorithm to implement the second set of fuzzy logic inference rules.
-
5. The method of claim 4, wherein the second fuzzy logic inference algorithm comprises:
-
defining a plurality of second fuzzy subsets, a second fuzzy subset being associated with each fuzzy logic inference rule of the second set of fuzzy logic inference rules;
forming a second fuzzy union set of the plurality of second fuzzy subsets; and
defuzzifying the second fuzzy union set by using a center of gravity defuzzification method to generate the second weight factor.
-
-
6. The method of claim 2, wherein the second set of fuzzy logic inference rules comprises:
-
b1)dP1(I) being a pixel value at a pixel I for the first difference signal of first order;
b2)dP2(I) being a pixel value at the pixel I for the second difference signal of first order; and
b3)K2 being the second weight factor;
wherein for the second set of fuzzy logic inference rules, the first difference signal of first order has grades of fuzzy set membership in fuzzy sets including Small (S) and Large (L);
wherein for the second set of fuzzy logic inference rules, the second difference signal of first order has grades of fuzzy set membership in fuzzy sets S and L; and
wherein for the second set of fuzzy logic inference rules, the second weight factor has grades of membership in fuzzy sets S and L.
-
-
7. The method of claim 6, wherein the second set of fuzzy logic inference rules comprises:
-
if dP1(I) is S, then K2 is L;
if dP1(I) is L, then K2 is S;
if dP2(I) is S, then K2 is L; and
if dP2(I) is L, then K2 is S.
-
-
8. The method of claim 1, wherein the first weighted sum at a pixel is characterized as:
(the first weight factor) * (the digital chrominance signal)+(1−
the first weight factor) * (the 2N-pixel delayed signal).
-
9. The method of claim 1, comprising:
generating the first weight factor by performing a first fuzzy logic inference algorithm to implement the first set of fuzzy logic inference rules.
-
10. The method of claim 9, wherein the first fuzzy logic inference algorithm comprises:
-
defining a plurality of first fuzzy subsets, a first fuzzy subset being associated with each fuzzy logic inference rule of the first set of fuzzy logic inference rules;
forming a first fuzzy union set of the plurality of first fuzzy subsets; and
defuzzifying the first fuzzy union set by using a center of gravity defuzzification method to generate the first weight factor.
-
-
11. The method of claim 1, wherein the first set of fuzzy logic inference rules comprises:
-
a1)dP1(I) being a pixel value at a pixel I for the first difference signal of first order;
a2)dP(I) being a pixel value at the pixel I for the difference signal of second order; and
a3)K1 being the first weight factor;
wherein for the first set of fuzzy logic inference rules, the first difference signal of first order has grades of fuzzy set membership in fuzzy sets including Negative Large (NL), Negative Small (NS), Positive Small (PS) and Positive Large (PL);
wherein for the first set of fuzzy logic inference rules, the difference signal of second order has grades of fuzzy membership in fuzzy sets NL, NS, PS and PL; and
wherein for the first set of fuzzy logic inference rules, the first weight factor has grades of fuzzy set membership in fuzzy sets Small (S) and Large (L).
-
-
12. The method of claim 11, wherein the first set of fuzzy logic inference rules comprises:
-
if dP1(I) is PL and if dP(I) is NL, then K1 is L;
if dP1(I) is PL and dP(I) is NS, then K1 is L;
if dP1(I) is PL and dP(I) is PS, then K1 is S;
if dP1(I) is PL and dP(I) is PL, then K1 is S;
if dP1(I) is PS and dP(I) is NL, then K1 is L;
if dP1(I) is PS and dP(I) is NS, then K1 is M;
if dP1(I) is PS and dP(I) is PS, then K1 is M;
if dP1(I) is PS and dP(I) is PL, then K1 is S;
if dP1(I) is NS and dP(I) is NL, then K1 is S;
if dP1(I) is NS and dP(I) is NS, then K1 is M;
if dP1(I) is NS and dP(I) is P, then K1 is M;
if dP1(I) is NS and dP(I) is PL, then K1 is L;
if dP1(I) is NL and dP(I) is NL, then K1 is S;
if dP1(I) is NL and dP(I) is NS, then K1 is S;
if dP1(I) is NL and dP(I) is PS, then K1 is L; and
if dP1(I) is NL and dP(I) is PL, then K1 is L.
-
-
13. A system of digital color transition improvement (DCTI) for a N-pixel delayed signal of a digital chrominance signal, the system comprising:
-
a first summing component for generating a first difference signal of first order using the digital chrominance signal and the N-pixel delayed signal;
a second summing component for generating a second difference signal of first order using the N-pixel delayed signal and a 2N-pixel delayed signal of the digital chrominance signal;
a third summing component for generating a difference signal of second order using the first and second difference signals of first order; and
a first calculation circuit block for calculating a first weighted sum of the digital chrominance signal and the 2N-pixel delayed signal, the fuzzy-logic circuit block being adapted to determine a first weight factor to characterize the first weighted sum by applying a first set of fuzzy logic inference rules to the first difference signal of first order and the difference signal of second order. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
a second calculation circuit block calculating a second weighted sum of the N-pixel delayed signal and the first weighted sum, the second weighted sum further characterized by a the second weight factor determined by applying a second set of fuzzy logic inference rules to the first and second difference signal of first order.
-
-
15. The system of claim 14, wherein the second weighted sum is characterized by:
(the second weight factor) * (the N-pixel delayed signal)+(1−
the second weight factor) * (the first weighted sum).
-
16. The system of claim 14, wherein the second fuzzy logic circuit block is adapted to generate the second weight factor by performing a second fuzzy logic inference algorithm that implements the second set of fuzzy logic inference rules.
-
17. The system of claim 16, wherein the second fuzzy logic inference algorithm is adapted to:
-
define a plurality of second fuzzy subsets, a second fuzzy subset being associated with each fuzzy logic inference rule of the second set of fuzzy logic inference rules;
form a second fuzzy union set of the plurality of second fuzzy subsets; and
defuzzify the second fuzzy union set by using a center of gravity defuzzification method to generate the second weight factor.
-
-
18. The system of claim 14, wherein the second set of fuzzy logic inference rules comprise:
-
b1) dP1(I) being a pixel value at a pixel I for the first difference signal of first order;
b2) dP2(I) being a pixel value at the pixel I for the difference signal of second order; and
b3) K2 being the second weight factor;
wherein for the second set of fuzzy logic inference rules, the first difference signal of first order has grades of fuzzy set membership in fuzzy sets Small (S) and Large (L);
wherein for the second set of fuzzy logic inference rules, the second difference signal of first order has grades of fuzzy membership in fuzzy sets S and L; and
wherein for the second set of fuzzy logic inference rules, the second weight factor has grades of membership in fuzzy sets S and L.
-
-
19. The system of claim 18, wherein the second set of fuzzy logic inference rules comprise:
-
if dP1(I) is S, then K2 is L;
if dP1(I) is L, then K2 is S;
if dP2(I) is S, then K2 is L; and
if dP2(I) is L, then K2 is S.
-
-
20. The system of claim 13, wherein the first weighted sum is characterized by:
(the first factor) * (the digital chrominance signal)+(1−
the first weight factor) * (the 2N-pixel delayed signal).
-
21. The system of claim 13, wherein the first fuzzy logic circuit block is adapted to generate the first weight factor by performing a first fuzzy logic inference algorithm that implements the first set of fuzzy logic inference rules.
-
22. The system of claim 21, wherein the first fuzzy logic inference algorithm is adapted to:
-
define a plurality of first fuzzy subsets, a first fuzzy subset being associated with each fuzzy logic inference rule of the first set of fuzzy logic inference rules;
form a first fuzzy union set of the plurality of first fuzzy subsets; and
defuzzify the first fuzzy union set by using a center of gravity defuzzification method to generate the first weight factor.
-
-
23. The system of claim 13, wherein the first set of fuzzy logic rules comprise:
-
a1) dP1(I) being a pixel value at a pixel I for the first difference signal of first order;
a2) dP(I) being a pixel value at pixel I for the difference signal of second order; and
a3) K1 being the first weight factor;
wherein for the first set of fuzzy logic inference rules, the first difference signal of first order has grades of fuzzy set membership in fuzzy sets including Negative Large (NL), Negative Small (NS), Positive Small (PS) and Positive Large (PL);
wherein for the first set of fuzzy logic inference rules, the difference signal of second order has grades of fuzzy membership in fuzzy sets NL, NS, PS and PL; and
wherein for the first set of fuzzy logic inference rules, the first weight factor has grades of fuzzy set membership in fuzzy sets Small (S) and Large (L).
-
-
24. The system of claim 23, wherein the first set of fuzzy logic inference rules comprise:
-
if dP1(I) is PL and if dP(I) is NL, then K1 is L;
if dP1(I) is PL and dP(I) is NS, then K1 is L;
if dP1(I) is PL and dP(I) is PS, then K1 is S;
if dP1(I) is PL and dP(I) is PL, then K1 is S;
if dP1(I) is PS and dP(I) is NL, then K1 is L;
if dP1(I) is PS and dP(I) is NS, then K1 is M;
if dP1(I) is PS and dP(I) is PS, then K1 is M;
if dP1(I) is PS and dP(I) is PL, then K1 is S;
if dP1(I) is NS and dP(I) is NL, then K1 is S;
if dP1(I) is NS and dP(I) is NS, then K1 is M;
if dP1(I) is NS and dP(I) is P, then K1 is M;
if dP1(I) is NS and dP(I) is PL, then K1 is L;
if dP1(I) is NL and dP(I) is NL, then K1 is S;
if dP1(I) is NL and dP(I) is NS, then K1 is S;
if dP1(I) is NL and dP(I) is PS, then K1 is L; and
if dP1(I) is NL and dP(I) is PL, then K1 is L.
-
-
25. A computer implemented method stored on a recordable media for improving color transition of a digital signal, comprising:
-
delaying the digital signal by N pixels to generate an N-pixel delayed signal;
delaying the N-pixel delayed signal by N pixels to generate a 2N-pixel delayed signal;
taking a difference of the digital signal and the N-pixel delayed signal to generate a first difference signal;
taking a difference of the N-delayed signal and the 2N-pixel delayed signal to generate a second difference signal;
taking a difference of the first and second difference signals to generate a second order difference signal;
generating a first weighted sum of the digital signal, the N-pixel delayed signal and a first weight factor, the first weight factor resulting from application of a first set of fuzzy logic inference rules; and
generating a second weighted sum of the first weighted sum, the N-pixel delayed signal and a second weight factor, the second weight factor resulting from application of a second set of fuzzy logic inference rules. - View Dependent Claims (26, 27, 28, 29, 30, 31, 32, 33)
generating the first weight factor by applying the first set of fuzzy logic inference rules to the second order difference signal and the first difference signal.
-
-
27. The method of claim 25 comprising:
generating the second weight factor by applying the second set of fuzzy logic inference rules to the first and second difference signals.
-
28. The method of claim 25 comprising:
-
defining a plurality of first fuzzy subsets, a first fuzzy subset corresponding to a fuzzy logic inference rule of the first set of fuzzy logic inference rules;
forming a first fuzzy union set of the plurality of first fuzzy subsets; and
defuzzifying the first fuzzy union set by using a center of gravity defuzzification method to generate the first weight factor.
-
-
29. The method of claim 28
wherein dP1(I) is a pixel value at a pixel I for the first difference signal; -
wherein dP(I) is a pixel value at the pixel I for the second difference signal;
wherein K1 is the first weight factor;
wherein for the first set of fuzzy logic inference rules, the first difference signal has grades of fuzzy set membership in fuzzy sets including Negative Large (NL), Negative Small (NS), Positive Small (PS) and Positive Large (PL);
wherein for the first set of fuzzy logic inference rules, the second order difference signal has grades of fuzzy membership in fuzzy sets NL, NS, PS and PL; and
wherein for the first set of fuzzy logic inference rules, the first weight factor has grades of fuzzy set membership in fuzzy sets Small (S) and Large (L).
-
-
30. The method of claim 29 wherein the first set of fuzzy logic inference rules comprises:
-
if dP1(I) is PL and if dP(I) is NL, then K1 is L;
if dP1(I) is PL and dP(I) is NS, then K1 is L;
if dP1(I) is PL and dP(I) is PS, then K1 is S;
if dP1(I) is PL and dP(I) is PL, then K1 is S;
if dP1(I) is PS and dP(I) is NL, then K1 is L;
if dP1(I) is PS and dP(I) is NS, then K1 is M;
if dP1(I) is PS and dP(I) is PS, then K1 is M;
if dP1(I) is PS and dP(I) is PL, then K1 is S;
if dP1(I) is NS and dP(I) is NL, then K1 is S;
if dP1(I) is NS and dP(I) is NS, then K1 is M;
if dP1(I) is NS and dP(I) is P, then K1 is M;
if dP1(I) is NS and dP(I) is PL, then K1 is L;
if dP1(I) is NL and dP(I) is NL, then K1 is S;
if dP1(I) is NL and dP(I) is NS, then K1 is S;
if dP1(I) is NL and dP(I) is PS, then K1 is L; and
if dP1(I) is NL and dP(I) is PL, then K1 is L.
-
-
31. The method of claim 25 comprising:
-
defining a plurality of second fuzzy subsets, a second fuzzy subset corresponding to a fuzzy logic inference rule of the second set of fuzzy logic inference rules;
forming a second fuzzy union set of the plurality of second fuzzy subsets; and
defuzzifying the second fuzzy union set by using a center of gravity defuzzification method to generate the second weight factor.
-
-
32. The method of claim 31
wherein dP1(I) is a pixel value at a pixel I for the first difference signal; -
wherein dP2(I) is a pixel value at the pixel I for the second difference signal;
wherein K2 is the second weight factor;
wherein for the second set of fuzzy logic inference rules, the first difference signal has grades of fuzzy set membership in fuzzy sets including Small (S) and Large (L);
wherein for the second set of fuzzy logic inference rules, the second difference signal has grades of fuzzy set membership in fuzzy sets S and L; and
wherein for the second set of fuzzy logic inference rules, the second weight factor has grades of membership in fuzzy sets S and L.
-
-
33. The method of claim 32 wherein the second set of fuzzy logic inference rules comprises:
-
if dP1(I) is S, then K2 is L;
if dP1(I) is L, then K2 is S;
if dP2(I) is S, then K2 is L; and
if dP2(I) is L, then K2 is S.
-
Specification