Packing and depacking histograms with statistical processing Original Research Article

A method is described and implemented which reduces the size of a histograms data base arising from a large set of histograms. The program compresses the histogram data before storage, avoiding significant loss of information and can be adapted to sample more densely areas of interest. The method has been applied to image processing where histograms were numbers of counts for the 256 gray levels of a subsampled image of millions of pixels.