Noise estimation based on a characteristic of digital video

1. A method of determining a characteristic of video data, the method comprising:

obtaining a set of N frames of the video data;

filtering the set of N frames using at least one filter to produce a set of N×T blocks of filtered video data, wherein T is a partition size associated with the at least one filter;

classifying each block in the set of N×T blocks as either a flat type block or a detailed type block;

associating with a frame in the set of N frames, a subset of blocks in the classified set of N×T blocks, wherein the subset of blocks are flat type blocks; and

determining the characteristic of video data based, at least in part, on the subset of blocks associated with the frame;

obtaining a next set of N frames of the video data; and

applying a noise reduction technique to the next set of N frames based on the characteristic of the video data.

2. The method of claim 1 , wherein the determining comprises:

determining m estimates of the characteristic of video data based on pixel values of blocks in the set of N×T blocks that have been classified as the flat type; and

selecting one of the m estimates as the characteristic of video data based, at least in part, on pixel values of the subset of blocks in the set of N×T blocks.

3. The method of claim 1 , wherein the subset of blocks comprises the set of N×T blocks.

4. The method of claim 1 , wherein the at least one filter comprises a diagonally oriented high-pass filter.

5. The method of claim 2 , wherein determining the m estimates of the characteristic of video data comprises:

partitioning the set of N frames of video data into m sets of N/m frames each; and

applying a different estimator to each of the m sets of N/m frames to produce a corresponding estimate of the characteristic of video data.

6. The method of claim 1 , wherein the characteristic of video data is a standard deviation value associated with noise present in the set of N frames.

7. The method of claim 1 , wherein the characteristic of video data is a median absolute deviation value associated with noise present in the set of N frames.

8. The method of claim 1 , further comprising:

inputting a set of T blocks of filtered video data to a maximum module to produce a first output, and

inputting the set of T blocks of filtered video data to a minimum module to produce a second output; and

determining at least one variance measure of a video frame in the set of N frames based on the first output and second output.

9. The method of claim 1 , further comprising:

inputting a set of T blocks of filtered video data to a maximum module to produce a first output, and

inputting the set of T blocks of filtered video data to a minimum module to produce a second output; and

determining at least one activity value associated with a video frame in the set of N frames based on the first output and second output.

10. A video processing system for determining a characteristic of video data, the video processing system comprising:

acquisition circuitry configured to:

obtain a set of N frames of the video data; and

obtain a next set of N frames of the video data;

filtering circuitry configured to apply at least one filter to the set of N frames to produce a set of N×T blocks of filtered video data, wherein T is a partition size associated with the at least one filter;

estimation circuitry configured to:

classify each block in the set of N×T blocks as either a flat type block or a detailed type block;

associate with a frame in the set of N frames, a subset of blocks in the classified set of N×T blocks, wherein the subset of blocks are flat type blocks; and

determine the characteristic of video data based, at least in part, on the subset of blocks associated with the frame; and

noise reduction circuitry configured to apply a noise reduction technique to the next set of N frames based on the characteristic of the video data.

11. The system of claim 10 , wherein the estimation circuitry is further configured to determine the characteristic of video data by:

determining m estimates of the characteristic of video data based on pixel values of blocks in the set of N×T blocks that have been classified as the flat type; and

selecting one of the m estimates as the characteristic of video data based, at least in part, on pixel values of the subset of blocks in the set of N×T blocks.

12. The system of claim 11 , wherein determining the m estimates of the characteristic of video data comprises:

partitioning the set of N frames of video data into m sets of N/m frames each; and

applying a different estimator to each of the m sets of N/m frames to produce a corresponding estimate of the characteristic of video data.

13. The system of claim 10 , wherein the subset of blocks comprises the set of N×T blocks.

14. The system of claim 10 , wherein the at least one filter comprises a diagonally oriented high-pass filter.

15. The system of claim 10 , wherein the characteristic of video data is a standard deviation value associated with noise present in the set of N frames.

16. The system of claim 10 , wherein the characteristic of video data is a median absolute deviation value associated with noise present in the set of N frames.

17. The system of claim 10 , wherein the estimation circuitry is further configured to:

input a set of T blocks of filtered video data to a maximum module to produce a first output, and

input the set of T blocks of filtered video data to a minimum module to produce a second output; and

determine at least one variance measure of a video frame in the set of N frames based on the first output and second output.

18. The system of claim 10 , wherein the estimation circuitry is further configured to:

input a set of T blocks of filtered video data to a maximum module to produce a first output, and

input the set of T blocks of filtered video data to a minimum module to produce a second output; and

determine at least one activity value associated with a video frame in the set of N frames based on the first output and second output.