METHOD FOR IDENTIFYING COLOR-BASED VECTORS FOR ANALYSIS OF FRAMES OF A VIDEO
A method for identifying color-based vectors is provided, for an improved analysis of frames of a video, for locally accurate color/material correspondence information is provided. The correspondence information can be used to improve the identification of illumination and material aspects of each image depicted in temporally spaced frames of a video.
1 . An automated, computerized method for analyzing multiple image files of a video, comprising the steps of:
providing at least two image files of the video, each depicting an image having image locations, in a computer memory;
selecting a first image file k and a second image file k-n, k-n being temporally spaced n frames from image file k in the video;
selecting an image location in one of the first image file and the second image file;
estimating a search location in the other of the first image file and the second image file;
searching the search location for an image location in the search location having a closest color match to the selected image location;
comparing a color match difference based upon the closest color match, to a threshold; and
computing a color flow vector, mapping the image location in the search location to the selected image location, when the color match difference is less than the threshold.
2 . The method of claim 1 including the additional step of using the color flow vector to define an anchor constraint between the selected image location and the image location in the search location, for use in an operation to segregate an image file into intrinsic illumination and material reflectance components.
3 . A computer program product, disposed on a non-transitory computer readable media, the product including computer executable process steps operable to control a computer to: receive at least two image files of a video, each depicting an image having image locations, select a first image file k and a second image file k-n, image file k-n being temporally spaced from image file k in the video, select an image location in one of the first image file and the second image file, estimate a search location in the other of the first image file and the second image file, search the search location for an image location in the search location having a closest color match to the selected image location, compare a color match difference based upon the closest color match, to a threshold and compute a color flow vector, mapping the image location in the search location to the selected image location, when the color match difference is less than the threshold.
4 . The computer program product of claim 3 including the additional step of using the color flow vector to define an anchor constraint between the selected image location and the image location in the search location, for use in an operation to segregate an image file into intrinsic illumination and material reflectance components.