Visual perception matching cost on binocular stereo images

An apparatus comprising:

• one or more processors;
• and
• one or more memories including computer program code, the one or more memories and the computer program code configured, with the one or more processors, to cause the apparatus to perform:
• taking a first image of a subject from a first viewpoint, said first image of said subject being composed of a first plurality of pixels;
• taking a second image of the subject from a second viewpoint, said second image of said subject being composed of a second plurality of pixels, wherein pixels in said first plurality of pixels in said first image correspond to pixels in said second plurality of pixels in said second image, said corresponding pixels lying on first epipolar lines in said first image and on second epipolar lines in said second image;
• rectifying said first and second images to make said first and second epipolar lines parallel to one another and to place corresponding pixels in said first and second images on one of a plurality of pairs of common scanlines;
• and
• for at least one pair of common scanlines, finding a corresponding pixel in said second image for each pixel in said first image such that a visual perception matching cost (VPMC) is minimized,
• wherein the visual perception matching cost (VPMC) is a function of:
• a local matching cost for the corresponding pixels on the pair of common scanlines in said first image and said second image;
• a minimum total difference sequence modeling how well the common scanlines match on one side of the corresponding pixels in said first image and said second image;
• and
• a minimum total difference sequence modeling how well the common scanlines match on the other side of the corresponding pixels in said first image and said second image.

A computer program product comprising a non-transitory computer-readable storage medium bearing computer program code embodied therein for use with an apparatus, the computer program code comprising code for performing at least the following:

• taking a first image of a subject from a first viewpoint, said first image of said subject being composed of a first plurality of pixels;
• taking a second image of the subject from a second viewpoint, said second image of said subject being composed of a second plurality of pixels, wherein pixels in said first plurality of pixels in said first image correspond to pixels in said second plurality of pixels in said second image, said corresponding pixels lying on first epipolar lines in said first image and on second epipolar lines in said second image;
• rectifying said first and second images to make said first and second epipolar lines parallel to one another and to place corresponding pixels in said first and second images on one of a plurality of pairs of common scanlines;
• and
• for at least one pair of common scanlines, finding a corresponding pixel in said second image for each pixel in said first image such that a visual perception matching cost (VPMC) is minimized,
• wherein the visual perception matching cost (VPMC) is a function of:
• a local matching cost for the corresponding pixels on the pair of common scanlines in said first image and said second image;
• a minimum total difference sequence modeling how well the common scanlines match on one side of the corresponding pixels in said first image and said second image;
• and
• a minimum total difference sequence modeling how well the common scanlines match on the other side of the corresponding pixels in said first image and said second image.