PROTECTION FILTER FOR IMAGE AND VIDEO PROCESSING

• US 20100061648A1
• Filed: 09/08/2008
• Published: 03/11/2010
• Est. Priority Date: 09/08/2008
• Status: Active Grant

• Current Assignee: ATI Technologies ULC (Advanced Micro Devices, Inc.)
• Original Assignee: ATI Technologies ULC (Advanced Micro Devices, Inc.)
• Inventors: Cao, Peter, Wong, Daniel
• Granted Patent: US 9,105,081 B2
• Time in Patent Office: Days
• US Class Current: 382/260
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/20182 Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering
• CPC Class Codes: G06T 2207/20201 Motion blur correction
• CPC Class Codes: G06T 5/20 using local operators
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 5/70 Denoising; Smoothing
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening

Abstract

A filter includes a conventional filtering block and a protection block. The conventional filtering block receives input values and provides filtered values. The protection block receives filtered values and a group of input values proximate the current input, to ensure that the output is lies within a range computed for the current input. The range is determined by the protection block based on the group of input values proximate the current input. Any algorithm or statistical function may be applied to the group of input values to determine the range. If a filtered value provided by the conventional filtering block is outside the range, then the protection block computes and outputs a value that is within the range. The filter may be used in temporal or spatial filtering of images and video to mitigate artifacts such as motion artifacts and static artifacts.

Claims

1. A digital filter comprising:

a filtering block receiving an input value from an input signal and providing a filtered value; and

a protection block comprising;

an input for receiving said filtered value, and a group of values proximate said input value in said input signal; and

control logic for computing minimum and maximum bounds based on said group of values; and

for forming an output value from said filtered value, such that said output value lies within the range of said bounds.

2. The filter of claim 1, wherein said control logic sets said output value to said maximum bound if said filtered value exceeds said maximum bound.

3. The filter of claim 1, wherein said control logic sets said output value to said minimum bound if said filtered value falls below said minimum bound.

4. The filter of claim 1, wherein said group of input values is proximate said input value in time within said input signal.

5. The filter of claim 1, wherein said group of input values is proximate said input value in space within said input signal.

6. The filter of claim 5, wherein said filter block receives a plurality of pixels comprising X_i,j[n], X_i,j[n−

1],X_i,j[n−

2] . . . corresponding to location (i,j) in a sequence of video frames F_n, F<sub>n−

1</sub>, F<sub>n−

2</sub>, . . . respectively; and

wherein said input value corresponds to X_i,j[n], and said group of input values proximate said input value comprise a set of pixels surrounding X_i,j[n] in F_n.

7. The filter of claim 6, wherein said group of input values is a P×

Q block of pixels { . . . ,X<sub>i−

1,j−

1</sub>[n], X<sub>i−

1,j</sub>[n], X<sub>i−

1,j+1</sub>[n], X<sub>i,j−

1</sub>[n], X_i,j[n], X_i,j+1[n], X<sub>i+1,j−

1</sub>[n], X_i+1,j[n], X_i+1,j+1[n], . . . } in F_n surrounding X_i,j[n].

8. The filter of claim 1, wherein said minimum bound is the smallest value in said group of values.

9. The filter of claim 1, wherein said maximum bound is the largest value in said group of values.

10. The filter of claim 1, wherein said group of values proximate said input value consists of M values, and said minimum bound is an average of the smallest N values in said group of values, for N<

M.

11. The filter of claim 1, wherein said group of values proximate said input value consists of M values, and said maximum bound is an average of the largest N values in said group of values, for N<

M.

12. A digital filter comprising:

a filter block receiving an input value from an input signal and providing a filtered value; and

a protection block comprising;

an input for receiving said filtered value, and a group of values proximate said input value in said input signal;

an output for providing an output value; and

control logic for computing a set of all possible output values corresponding to said input value based on said group of values, said logic forming and providing said output value by selecting an element of said set.

13. A spatial filter comprising:

a scaling block for scaling an input image comprising input pixels, to provide a scaled image comprising scaled pixels;

said scaling block providing a scaled pixel corresponding to at least one input pixel;

a protection block comprising;

an input for receiving said scaled pixel, and a group of pixels proximate said at least one pixel in said input image; and

a control logic to compute minimum and maximum bounds for said scaled pixel based on said group of pixels and to provide an output value corresponding to said scaled pixel such that said output value lies within the range of said bounds.

14. The filter of claim 13, wherein said scaling block accepts a second group of pixels proximate said at least one input pixel to provide said scaled pixel, and said control logic sets said output value to said maximum bound if said scaled pixel exceeds said maximum bound.

15. The filter of claim 13, wherein said scaling block accepts a second group of pixels proximate said at least one input pixel to provide said scaled pixel, and said control logic sets said output value to said minimum bound if said scaled pixel falls below said minimum bound.

16. A method of providing an output corresponding to a current input value in an input signal, said method comprising:

i) receiving said input value;

ii) receiving a group of input values proximate said input value in said input signal;

iii) filtering said input value to provide a filtered value;

iv) computing a minimum bound and a maximum bound for said filtered value based on said group of input values proximate said input value in said input signal; and

v) forming said output value from said filtered value such that said output value lies within the range of said bounds, and providing said output value.

17. The method of claim 16, wherein said output value is set to said maximum bound if said filtered value exceeds said maximum bound.

18. The method of claim 16, wherein said output value is set to said minimum bound if said filtered value falls below said minimum bound.

19. The method of claim 16, wherein said group of input values is proximate said input value in time, in said input signal.

20. The method of claim 16, wherein said group of input values is proximate said input value in space, in said input signal.

21. The method of claim 16, wherein said input signal comprises a sequence of input values X_i,j[n], X_i,j[n−

1], X_i,j[n−

2] . . . corresponding to pixels at location (i,j) in a sequence of video frames F_n, F<sub>n−

1</sub>, F<sub>n−

2</sub> . . . ; and

said input value corresponds to X_i,j[n].

22. A device comprising a processor in communication with memory, said memory storing processor executable instructions causing said processor to:

i) receive a current input value in an input signal;

ii) receive a group of input values proximate said input value in said input signal;

iii) filter said input value to provide a filtered value;

iv) compute a minimum bound and a maximum bound for said filtered value based on said group of input values proximate said input value in said input signal; and

v) form an output value from said filtered value such that said output value lies within the range of said bounds; and

vi) provide said output value.

23. The device of claim 22, wherein said processor comprises one of a digital signal processor (DSP), a central processing unit (CPU) and a graphical processing unit (GPU).

24. The device of claim 22, wherein said group of input values is proximate said input value in time, in said input signal.

25. The device of claim 22, wherein said group of input values is proximate said input value in space, in said input signal.