Systems and methods for generating a virtual camera viewpoint for an image
A color image and a depth image of a live video are received. A user is extracted from the information of the color image and the depth image. Spurious depth vales may be corrected. Points or pixels of an image as seen from a viewpoint of a reference camera at a reference camera location are mapped to points of the image as would be seen from a viewpoint of a virtual camera at a virtual camera location. As such, a transformed color image is generated. Disoccluded pixels may be processed to address any gaps within the transformed color image.
1. A method for generating a transformed image, the method comprising:
receiving, using a computer, a depth image and an input color image of a video from at least one reference camera at a reference camera location, the input color image comprising pixels arranged in rows and columns;
receiving input identifying a virtual camera location; and
generating, using the computer, a transformed color image reflecting a viewpoint from the virtual camera location, wherein the generating includes a three dimensional (3D) mapping of the pixels of the input color image to pixels of the transformed color image, wherein the mapping comprises reversing a projection matrix of the at least one reference camera based on depth value—s—of the pixel—s—from the depth image and on the virtual camera location.
2. The method as set forth in claim 1 , further comprising correcting spurious depth values of the depth image, the correcting comprising the application of a bilateral filter to regions of the depth image that comprise an unknown depth value.
3. The method as set forth in claim 1 , further comprising processing disoccluded pixels of portions of the transformed color image that were not captured by the reference camera, the processing comprising assigning a weighted average pixel color value of nearby pixels of the transformed color image to each disoccluded pixel in the transformed color image.
4. The method as set forth in claim 1 , further comprising processing disoccluded pixels of portions of the transformed color image that were not captured by the reference camera, the processing comprising filling in a region comprising disoccluded pixels with an appearance of a shadow such that a light source associated with the shadow corresponds with the reference camera location.
5. A system comprising at least one processor and non-transitory memory, the memory storing instructions that, when executed by the one or more processors, are operative:
to receive a depth image and an input color image of a video from at least one reference camera at a reference camera location, the input color image comprising pixels arranged in rows and columns;
to receive input identifying a virtual camera location; and
to generate a transformed color image reflecting a viewpoint from the virtual camera location, wherein the generating includes a three dimensional (3D) mapping of the pixels of the input color image to pixels of the transformed color image, wherein the mapping comprises reversing a projection matrix of the at least one reference camera based on depth value—s—of the pixel—s—from the depth image and on the virtual camera location.
6. The system as set forth in claim 5 , further comprising a module for correcting spurious depth values of the depth image, the correcting comprising the application of a bilateral filter to regions of the depth image that comprise an unknown depth value.
7. The system as set forth in claim 5 , further comprising a module for processing disoccluded pixels of portions of the transformed color image that were not captured by the reference camera, the processing comprising assigning a weighted average pixel color value of nearby pixels of the transformed color image to each disoccluded pixel in the transformed color image.
8. The system as set forth in claim 5 , further comprising a module for processing disoccluded pixels portions of the transformed color image that were not captured by the reference camera, the processing comprising filling in a region comprising disoccluded pixels with an appearance of a shadow such that a light source associated with the shadow corresponds with the reference camera location.
9. A non-transitory computer readable medium containing one or more instructions, wherein the one or more instructions, when executed by one or more processors, cause the one or more processors to perform the steps of:
receiving, using a computer, a depth image and an input color image of a video from at least one reference camera at a reference camera location;
receiving input identifying a virtual camera location; and
generating, using the computer, a transformed color image reflecting a viewpoint from the virtual camera location, wherein the generating includes a three dimensional (3D) mapping of the pixels of the input color image to pixels of the transformed color image, wherein the mapping comprises reversing a projection matrix of the at least one reference camera based on depth value—s—of the pixel—s—from the depth image and on the virtual camera location.
10. The non-transitory computer readable medium as set forth in claim 9 , wherein the steps further comprise correcting spurious depth values of the depth image the correcting comprising the application of a bilateral filter to regions of the depth image that comprise an unknown depth value.
11. The non-transitory computer readable medium as set forth in claim 9 , wherein the steps further comprise processing disoccluded pixels of portions of the transformed color image that were not captured by the reference camera, the processing comprising assigning a weighted average pixel color value of nearby pixels of the transformed color image to each disoccluded pixel in the transformed color image.
12. The non-transitory computer readable medium as set forth in claim 9 , wherein the steps further comprise processing disoccluded pixels of portions of the transformed color image that were not captured by the reference camera, the processing comprising filling in a region comprising disoccluded pixels with an appearance of a shadow such that a light source associated with the shadow corresponds with the reference camera location.
13. The method of claim 1 , wherein the virtual camera location is displaced vertically from the reference camera location.
14. The method of claim 1 , wherein the virtual camera location is displaced horizontally from the reference camera location.
15. The method of claim 1 , wherein the virtual camera location is displaced both horizontally and vertically from the reference camera location.
16. The system of claim 5 , wherein the virtual camera location is displaced vertically from the reference camera location.
17. The system of claim 5 , wherein the virtual camera location is displaced horizontally from the reference camera location.
18. The system of claim 5 , wherein the virtual camera location is displaced both horizontally and vertically from the reference camera location.
19. The non-transitory computer readable medium of claim 9 , wherein the virtual camera location is displaced vertically from the reference camera location.
20. The non-transitory computer readable medium of claim 9 , wherein the virtual camera location is displaced horizontally from the reference camera location.
21. The non-transitory computer readable medium of claim 9 , wherein the virtual camera location is displaced both horizontally and vertically from the reference camera location.