Image data clustering and color conversion

Abstract

Methods and systems herein provide for color conversion of image data to another color space. Such color conversion includes converting image data to a color space by identifying color regions in the image data. One color conversion system herein includes a color region identifier operable to select color values in the image data, determine numerical centers of the selected color values, and generate color regions, or color value “clusters”, based on the selected color values and the numerical centers. The color conversion system also includes an optimization module operable to refine the numerical center of each color region, determine a boundary for each color region based on the refined center, and remove a portion of the color cluster centers to optimize a number of the color regions for color conversion processing.

1 . A color conversion system operable to identify color regions in image data and convert the image data to a color space, the system comprising:
 a color region identifier operable to select color values in the image data, determine numerical centers of the selected color values, and generate color regions based on the selected color values and the numerical centers; and   an optimization module operable to refine the numerical center of each color region, determine a boundary for each color region based on the refined center, and remove a portion of the numerical centers to optimize a number of the color regions for color conversion processing.   
     
     
         2 . The color conversion system of  claim 1 , wherein the optimization module is further operable to compute a covariance matrix of a color region to optimize the number of the color regions. 
     
     
         3 . The color conversion system of  claim 2 , wherein the optimization module is further operable to remove the color region from subsequent processing based on a rank of the covariance matrix of the color region. 
     
     
         4 . The color conversion system of  claim 3 , wherein the optimization module is further operable to configure the color values of the removed color region with an adjacent color region. 
     
     
         5 . The color conversion system of  claim 1 , wherein the color region identifier is further operable to randomly select the numerical center of the selected color values when determining the numerical center of the selected color values. 
     
     
         6 . The color conversion system of  claim 1 , wherein the optimization module is further operable to determine a Voronoi set of the selected color values, iteratively update the Voronoi set, and refine the numerical center of the selected color values via an orthogonal least squares algorithm. 
     
     
         7 . The color conversion system of  claim 1 , further comprising a converter operable to generate a numerical model of the color region using a radial basis function and convert the color region to a color space of an imaging device. 
     
     
         8 . The color conversion system of  claim 1 , wherein the optimization module is further operable to determine a boundary for each color region based on human perceptions properties. 
     
     
         9 . A method of determining color regions in image data, the method comprising:
 selecting clusters of color values in the image data;   determining random center values of the clusters;   iteratively determining boundaries of the clusters to define the color regions; and   processing subset color values of the clusters to remove clusters from subsequent processing.   
     
     
         10 . The method of  claim 9 , wherein processing subset color values comprises:
 generating a covariance matrix of each cluster;   determining a rank of each covariance matrix; and   analyzing the rank of each covariance matrix to determine subsequent processing of a respective cluster and optimize a number of the clusters.   
     
     
         11 . The method of  claim 10 , further comprising:
 moving color values of a removed cluster to an adjacent cluster in response to analyzing the rank of the covariance matrix of the removed cluster; and   repeating said processing subset color values of the clusters until each cluster includes a full rank covariance matrix.   
     
     
         12 . The method of  claim 9 , wherein determining the random center values of the clusters comprises:
 generating a Voronoi set of the selected color values;   iteratively updating the Voronoi set; and   refining the center values of the clusters via an orthogonal least squares algorithm.   
     
     
         13 . The method of  claim 9 , further comprising:
 generating numerical models of the color regions using radial basis functions; and   converting the color regions to a color space based on the numerical models.   
     
     
         14 . The method of  claim 9 , wherein iteratively determining boundaries of the clusters to define the color regions is based at least in part on human perceptions properties. 
     
     
         15 . A printing system, comprising:
 a printer operable to print to a tangible medium; and   a printer controller operable to process a print job from a host system and transfer the processed print job to the printer, wherein the print job includes image data having color values and wherein the printer controller comprises:   a color region identifier operable to select the color values in the image data, determine numerical centers of the selected color values, and generate color regions based on the selected color values and the numerical centers;   an optimization module operable to refine the center of the color region and determine a boundary for each color region based on the refined center and remove a portion of the numerical centers to optimize a number of the color regions for color conversion processing; and   a converter operable to convert the color regions to a color space prior to printing.   
     
     
         16 . A color conversion system operable to identify color regions in image data and convert the image data to a color space, the system comprising:
 a color region identifier operable to select clusters of color values in the image data and determine random center values of the clusters; and   an optimization module operable to iteratively determine boundaries of the clusters to define the color regions and process subset color values of the clusters to remove clusters from subsequent processing.   
     
     
         17 . The color conversion system of  claim 16 , wherein the optimization module is further operable to generate a covariance matrix of each cluster, determine a rank of each covariance matrix, and analyze the rank of each covariance matrix to determine subsequent processing of a respective cluster and optimize a number of the clusters. 
     
     
         18 . The color conversion system of  claim 17 , wherein the optimization module is further operable to move color values of a removed cluster to an adjacent cluster in response to analyzing the rank of the covariance matrix of the removed cluster and process subset color values of the clusters until each cluster includes a full rank covariance matrix. 
     
     
         19 . The color conversion system of  claim 16 , wherein the color region identifier is further operable to generate a Voronoi set of the selected color values, iteratively update the Voronoi set, and refine the center values of the clusters via an orthogonal least squares algorithm. 
     
     
         20 . The color conversion system of  claim 16 , further comprising a converter operable to generate a numerical model of the color regions using radial basis functions and to convert the color regions to a color space of an imaging device based on the numerical models.