Interactive tone mapping for high dynamic range video

1. A method comprising the steps of:

selecting a frame of a video sequence comprising a plurality of frames;

providing a user interface for a user to manually apply scribble marks to the frame;

manually applying N scribble marks to the frame, N being a whole number of value 1 or greater; and

tone mapping multiple frames in the video sequence responsive to the scribble marks.

2. The method of claim 1 comprising:

classifying pixels in the frame to N categories, the N categories correlate to the N scribble marks.

3. The method of claim 1 comprising:

classifying pixels in the frame to N categories, the N categories correlate to the N scribble marks, wherein the classifying is applied to all frames in the video sequence.

4. The method of claim 3 comprising:

generating an initial exposure map for the frame, wherein the initial exposure map is used to classify the pixel in the classifying step.

5. The method of claim 4 comprising:

applying an edge preserving filter to generate a final exposure map, wherein the tone mapping step is further performed responsive to the final exposure map.

6. The method of claim 3 comprising:

creating feature vectors from initial input pixel data of the video sequence for each pixel along the N scribe marks, wherein classifying step is responsive to the feature vectors.

7. The method of claim 6 comprising:

applying at least one Gaussian mixture model in creating the feature vectors.

8. The method of claim 6 comprising:

applying at least one Gaussian mixture model in creating the feature vectors, wherein the feature vectors of the pixels at least include three luminance values, a first value for the luminance of a given pixel itself, and the others being average luminance of the neighboring pixels using different window sizes.

9. The method of claim 6 comprising:

applying an edge preserving filter to generate a final exposure map for each frame, wherein the tone mapping is performed by applying the final exposure map for each frame.

10. The method of claim 4 comprising:

applying an edge preserving filter to generate a final exposure map, wherein the edge preserving filter is a bilateral filter or a weight least square filter;

wherein the tone mapping is performed by applying the final exposure map for each frame.

11. The method of claim 1 , wherein the N scribble marks is non-continuous.

12. A device, comprising:

an interface configured to accept entry of a selected frame of a video sequence comprising a plurality of frames and to accept N manually applied scribble marks on the frame by a user input, N being a whole number of value 1 or greater; and

a processor configured to tone map multiple frames in the video sequence responsive to the scribble marks.

13. The device of claim 12 , wherein the processor is further configured to classify pixels in the frame to N categories, the N categories correlate to the N scribble marks.

14. The device of claim 12 , wherein the processor is further configured to classify pixels in the frame to N categories, the N categories correlate to the N scribble marks, wherein the classification is applied to all frames in the video sequence.

15. The device of claim 14 , wherein the processor is further configured to generate an initial exposure map for the frame, wherein the initial exposure map is used to classify the pixel in the classifying step.

16. The device of claim 15 , wherein the processor is further configured to apply an edge preserving filter to generate a final exposure map, wherein the processor is configured to tone map responsive to the final exposure map.

17. The device of claim 14 , wherein the processor is further configured to create feature vectors from initial input pixel data of the video sequence for each pixel along the N scribe marks, wherein the processor is configured to classify responsive to the feature vectors.

18. The device of claim 17 , wherein the processor is further configured to apply at least one Gaussian mixture model in creating the feature vectors.

19. The device of claim 17 , wherein the processor is further configured to apply at least one Gaussian mixture model in creating the feature vectors, wherein the feature vectors of the pixels at least include three luminance values, a first value for the luminance of a given pixel itself, and the others being average luminance of the neighboring pixels using different window sizes.

20. The device of claim 17 , wherein the processor is further configured to apply an edge preserving filter to generate a final exposure map for each frame, wherein the processor is configured to tone map by applying the final exposure map for each frame.

21. The device of claim 15 , wherein the processor is further configured to apply an edge preserving filter to generate a final exposure map, wherein the edge preserving filter is a bilateral filter or a weight least square filter; wherein the tone mapping is performed by applying the final exposure map for each frame.

22. The device of claim 12 , wherein the N scribble marks are non-continuous.