Method and apparatus for pre-processing mixed raster content planes to improve the quality of a decompressed image and increase document compression ratios

A method of processing a pixel map representation of a document, comprising the steps of:

• a) obtaining a set of image data signals which represent light intensity in a portion of a digital image;
• b) creating a selector plane from said image data signals which separates said signals into subsets;
• c) creating a first image plane using said selector plane and said image data signals, wherein said first image plane contains a first subset of image data signals;
• d) creating a second image plane using said selector plane and said image data signals, wherein said second image plane contains a second subset of image data signals;
• and
• e) processing at least one of said first image plane and said second image plane based upon data contained in said selector plane, wherein said processing step further comprises the steps of:
• inputting a threshold signal, which indicates an acceptable level of distortion for a subsequent processing operation;
• inputting a variance signal, which indicates an average magnitude between an image data signal and an average value of all signals in said subset;
• generating an image plane map which identifies locations in an associated image data subset in which image data signals that shall be used to reconstruct said digital image reside;
• and
• replacing all image data signals on said image plane map that correspond to associated subset locations not identified for image reconstruction with image data signals which have values that will enhance said image reconstruction and outputting said subset if said variance signal is greater than or equal to said threshold signal and at least one, but fewer than all locations in said image plane map associated with said subset have been identified for said image reconstruction.