Detection of Gaussian noise in video signals
First Claim
Patent Images
1. A method of detecting Gaussian noise in a video signal comprising the steps of:
- finding best qualified blocks in an image of the video signal that are substantially uniform in luminance;
calculating an average standard deviation for the best qualified blocks;
temporally filtering the average standard deviation over other images of the video signal to obtain a smoothed average standard deviation; and
calibrating the smoothed average standard deviation into a scale of corresponding input noise levels to obtain the Gaussian noise in the video signal.
4 Assignments
0 Petitions
Accused Products
Abstract
A method and apparatus for the detection of Gaussian noise in a video signal decomposes an image from the video signal to obtain best qualified blocks having a relatively uniform luminance. From the best qualified blocks an average standard deviation is calculated, and then smoothed by temporal filtering. The filtered average standard deviation is finally calibrated against a scale of corresponding input noise levels to obtain the Gaussian noise in the video signal.
-
Citations
4 Claims
-
1. A method of detecting Gaussian noise in a video signal comprising the steps of:
-
finding best qualified blocks in an image of the video signal that are substantially uniform in luminance;
calculating an average standard deviation for the best qualified blocks;
temporally filtering the average standard deviation over other images of the video signal to obtain a smoothed average standard deviation; and
calibrating the smoothed average standard deviation into a scale of corresponding input noise levels to obtain the Gaussian noise in the video signal. - View Dependent Claims (2, 3, 4)
subjecting a luminance bit-map of an image to a spatial decomposition to find a significant number of flat blocks; and
applying local variance and local luminance criteria to the flat blocks to chose a set of representative flat blocks as the best qualified blocks.
-
-
3. The method as recited in claim 2 wherein the finding step further comprises the step of selecting from the representative flat blocks a subset of representative flat blocks as the best qualified blocks.
-
4. The method as recited in claims 1, 2 or 3 wherein the best qualified blocks have an overall size no smaller than a specified minimum.
Specification