Apparatus and Method for Effect Pigment Identification

Abstract

A computer-implemented method for identifying an effect pigment, the method comprising executing, on at least one processor of at least one computer, steps of: a) acquiring sample image data describing a digital image of a layer comprising a sample effect pigment b) determining, based on the sample image data, sparkle point data describing a sample distribution of sparkle points defined by the digital image, wherein the sample distribution is defined in an N-dimensional color space, wherein N is an integer value equal to or larger than 3; c) determining, based on the sparkle point data, sparkle point transformation data describing a transformation of the sample distribution into an (N-1)-dimensional color space; d) determining, based on the sparkle point transformation data, sparkle point distribution geometry data describing a geometry of the sample distribution; e) acquiring reference distribution geometry data describing a geometry of a reference distribution of sparkle points in the (N-1)-dimensional color space; f) acquiring reference distribution association data describing an association between the reference distribution and an identifier of the reference distribution; g) determining, based on the sparkle point distribution geometry data and the reference distribution geometry data and the reference distribution association data, sample pigment identity data describing an identity of the sample effect pigment.

1 - 22 . (canceled) 
     
     
         23 . A computer-implemented method for characterizing effect pigment spots in images of a layer, the method comprising executing, on at least one processor of at least one computer, the steps of:
 a) forming one or more point clouds of effect pigment spots in a three-dimensional color space, each point in the point clouds being characterized by three device-independent color coordinates in the three-dimensional color space, the device-independent color coordinates being:
 i) a first and a second linearly projected chromaticity coordinates, and 
 ii) a lightness-related value; 
   b) computing a distance of the one or more point clouds to one or more reference point clouds in the three-dimensional color space, wherein the one or more reference point clouds are formed using the same color coordinates as the one or more point clouds of effect pigment spots, and wherein each reference point cloud is associated with a reference effect pigment;   c) selecting, for each of the one or more point clouds, the reference point cloud with shortest distance; and   d) characterizing the effect pigment spots in the one or more point clouds as belonging to the reference effect pigment associated with the respective selected reference point cloud.   
     
     
         24 . The method of  claim 23 , wherein computing the distance comprises at least one of the steps of:
 encoding the one or more point clouds into a hierarchical tree data structure;   forming a point cloud envelope;   computing the cloud’s centroid projected on a plane of a chromaticity diagram; and   computing a measure of the cloud’s statistical dispersion.   
     
     
         25 . The method of  claim 23 , wherein forming the one or more point clouds comprises the steps of:
 computing the lightness of one or more effect pigment spots within one or more images;   computing a lightness mean of effect pigment spots; and   computing a deviation from the lightness mean of the effect pigment spots,   wherein the lightness-related value of the points in the point cloud equals the deviation from the lightness mean.   
     
     
         26 . The method of  claim 23 , comprising correcting the value of one or more of the device-independent color coordinates by a color value measured over a surface the size of which corresponds to that of at least 9 effect pigment spots. 
     
     
         27 . The method of  claim 23 , comprising applying a thresholding method to the images of the layer to form images comprising effect pigment spots. 
     
     
         28 . The method of  claim 23 , wherein forming one or more point clouds comprises forming one or more point cloud clusters and/or executing a point clustering method. 
     
     
         29 . The method of  claim 23 , wherein the method further comprises a step of comparing the point cloud to one or more reference point clouds, in particular, using a two- or more-dimensional pattern matching method. 
     
     
         30 . The method of  claim 23 , wherein the images are acquired at a plurality of viewing and illumination angle combinations. 
     
     
         31 . The method of  claim 30 , wherein the method further comprises a step of automatically comparing the plurality of point clouds acquired at a plurality of viewing and illumination angle combinations to one or more reference point clouds with a three-dimensional pattern matching method. 
     
     
         32 . The method of  claim 23 , comprising displaying the point clouds on a plot comprising three or more axes, wherein a first axis is for the first chromaticity coordinate u′, a second axis is for the second chromaticity coordinate v′, and a third axis is for the deviation from lightness mean (L* - <L*> ) of the effect pigment spots. 
     
     
         33 . The method of  claim 23 , further comprising:
 providing one or more digital cameras;   providing one or more illumination sources;   providing one or more electronic displays;   providing a plurality of reference point clouds stored in non-volatile semiconductor memory; and   acquiring color image data from the digital cameras.   
     
     
         34 . A method for characterizing effect pigment spots in images of a layer, the method comprising the steps of:
 a) forming one or more point clouds of effect pigment spots in a three-dimensional color space, each point in the point clouds being characterized by three device-independent color coordinates in the three-dimensional color space, the device-independent color coordinates being:
 i) a first and a second linearly projected chromaticity coordinates, and 
 ii) a lightness-related value; 
   b) analyzing the one or more point clouds using a classifier, wherein the classifier has been trained on a plurality of reference point clouds of reference effect pigment spots to compute a distance of the one or more point clouds to each of said reference point clouds of reference effect pigment spots, wherein the one or more reference point clouds are formed in the three-dimensional color space using the same color coordinates as the one or more point clouds of effect pigment spots, and wherein each reference point cloud is associated with a reference effect pigment; and   c) based on a result of the step of analyzing, outputting output data comprising one or more of:   i) a pigment identity of one or more effect pigments corresponding to the one or more of the point cloud clusters, 
 ii) an effect pigment class related to the material or morphology of the pigment or pigment flakes, or 
 iii) a pigment grade or coarseness of each of the effect pigments identified in the layer. 
   
     
     
         35 . The method of  claim 34 , further comprising forming one or more point cloud clusters from the one or more point clouds of effect pigment spots in the three-dimensional color space, wherein the step of analyzing is carried out for the one or more point cloud clusters. 
     
     
         36 . The method of  claim 34 , wherein computing the distance comprises at least one of the steps of:
 encoding the one or more point clouds into a hierarchical tree data structure;   forming a point cloud envelope;   computing the cloud’s centroid projected on a plane of a chromaticity diagram; and   computing a measure of the cloud’s statistical dispersion.   
     
     
         37 . The method of  claim 34 , wherein forming the one or more point clouds comprises the steps of:
 computing the lightness of one or more effect pigment spots within one or more images;   computing a lightness mean of effect pigment spots; and   computing a deviation from lightness mean of the effect pigment spots,   wherein the lightness-related value of the points in the point cloud equals the deviation from lightness mean.   
     
     
         38 . The method of  claim 34 , wherein the images are acquired at a plurality of viewing and illumination angle combinations. 
     
     
         39 . The method of  claim 38 , wherein the method comprises automatically comparing a plurality of point clouds acquired at a plurality of viewing and illumination angle combinations to said one or more reference point clouds. 
     
     
         40 . A computer-implemented method for characterizing effect pigment spots in images of a layer, the method comprising executing, on at least one processor of at least one computer, the steps of:
 a) forming one or more point clouds of effect pigment spots in a three-dimensional color space, each point in the point clouds being characterized by three device-independent color coordinates in the three-dimensional color space, the device-independent color coordinates being a first and a second linearly projected chromaticity coordinates and a lightness-related value, by
 i) computing the lightness of one or more effect pigment spots within one or more images; 
 ii) computing a lightness mean of effect pigment spots; and 
 iii) computing a deviation from the lightness mean of the effect pigment spots, wherein the lightness-related value of the points in the point cloud equals the deviation from the lightness mean; 
   b) computing a distance of the one or more point clouds to one or more reference point clouds in the three-dimensional color space by
 i) encoding the one or more point clouds into a hierarchical tree data structure; 
 ii) forming a point cloud envelope; 
 iii) computing the cloud’s centroid projected on a plane of a chromaticity diagram; and 
 iv) computing a measure of the cloud’s statistical dispersion, 
 wherein the one or more reference point clouds are formed using the same color coordinates as the one or more point clouds of effect pigment spots, and wherein each reference point cloud is associated with a reference effect pigment;   c) selecting, for each of the one or more point clouds, the reference point cloud with shortest distance; and   d) characterizing the effect pigment spots in the one or more point clouds as belonging to the reference effect pigment associated with the respective selected reference point cloud.