초록 ▼

A method for operating a data processing system to generate a second image from a first image having partially sampled color values at each pixel. The first image includes a two-dimensional array of pixel values, each of the pixel values corresponding to the light intensity in one of a plurality of

A method for operating a data processing system to generate a second image from a first image having partially sampled color values at each pixel. The first image includes a two-dimensional array of pixel values, each of the pixel values corresponding to the light intensity in one of a plurality of spectral bands at a location in the first image. The second image includes a second two-dimensional array of color vectors. Each color vector has a light intensity value for each of the spectral bands. There is one such vector corresponding to each location having a pixel value in the first image. One component of the vector is equal to the pixel value in the first image at that location. The present invention computes the missing color components at each location. The method begins by providing an estimate for each component that is not equal to one of the pixel values from the first image for each vector. An updated estimate for each of the estimates is then generated by optimizing a weighted sum of first and second functions. The first function measures the degree of roughness of the second image, and the second function measures the degree to which the vectors change direction between neighboring locations in the second image.

대표청구항 ▼

1. A method for operating a data processing system to generate a second image from a first image, said first image comprising a two-dimensional array of pixel values, each of said pixel values corresponding to the light intensity in one of a plurality of spectral bands at a location in said first im

1. A method for operating a data processing system to generate a second image from a first image, said first image comprising a two-dimensional array of pixel values, each of said pixel values corresponding to the light intensity in one of a plurality of spectral bands at a location in said first image, and said second image comprising a second two-dimensional array of color vectors, each of said vectors having a light intensity value for each of said spectral bands, there being one such vector corresponding to each location having a pixel value in said first image, one component of said vector being equal to said pixel value in said first image at that location, said method comprising the steps of: providing an estimate for each component that is not equal to one of said pixel values from said first image for each vector; and generating an updated estimate for each of said estimates by optimizing a weighted sum of a first function that measures the degree of roughness of said second image and a second function that measures the degree to which said vectors change direction between neighboring locations in said second image. 2. The method of claim 1 wherein said step of generating an updated estimate comprises optimizing said weighted sum using a Gauss-Seidel iterative method. 3. The method of claim 1 wherein said second function comprises a sum of the magnitude of the vector product of each of said vectors adjacent to that vector in said second two-dimensional array.