Temporal stability for single frame super resolution
A method includes receiving, at a processor, at least one frame of input image data, producing motion vector fields between the frames of input image data, and applying temporal stability to the at least one frame of the input image data to produce noise reduced image data, wherein applying temporal stability comprises separating pixel data into frequency bands.
1. A method, comprising:
receiving, at a processor, at least one frame of input image data;
producing motion vector fields between the frames of input image data; and
applying temporal stability to the at least one frame of the input image data to produce noise reduced image data, wherein applying temporal stability comprises separating pixel data into frequency bands.
2. The method of claim 1 , wherein separating the pixel data into frequency bands comprises applying a filter to the pixel data.
3. The method of claim 2 , wherein the filter comprises a bilateral filter.
4. The method of claim 1 , wherein applying temporal stability comprises filtering a frame of input image data based upon motion errors in high and low frequency data.
5. The method of claim 4 , wherein the separating the motion errors into bands is based on error measurement ratios.
6. The method of claim 5 wherein filtering the frame of input image data comprises applying one of either a finite impulse response filter, or an infinite impulse response filter.
7. The method of claim 1 , wherein applying temporal stability is constrained by a spatial noise reduction process.
8. The method of claim 1 , wherein more than one noise reduced pixel is calculated for each output pixel using more than one motion vector, and a final noise reduced pixel is calculated using a combination of the more than one noise reduced pixels.
9. The method of claim 7 wherein the spatial noise reduction process uses a bilateral filter.
10. A method, comprising:
receiving, at a processor, at least one frame of input image data;
producing motion vector fields between the frames of input image data;
applying temporal stability to the at least one frame of the input image data to produce noise reduced image data, wherein applying temporal stability comprises filtering a frame of input image data based upon motion errors in high and low frequency data, and separating pixel data into frequency bands based upon motion errors in the high and the low frequency data and error measurement ratios.
11. The method of claim 10 , wherein filtering the frame of input image data comprises applying one of either a finite impulse response filter or an infinite impulse response filter.
12. The method of claim 10 , wherein the error measurement ratio is a ratio between the sum of absolute differences for the high frequency data and the low frequency data.
13. The method of claim 12 , wherein the ratio is based upon an amount of texture in a neighborhood of the pixel.
14. A method, comprising:
receiving, at a processor, at least one frame of input image data;
producing motion vector fields between the frames of input image data; and
applying temporal stability to the at least one frame of the input image data to produce noise reduced image data, wherein applying temporal stability comprises filtering a frame of input image data based upon motion errors in high and low frequency data and separating pixel data into frequency bands.