Fast low-memory paper color suppression algorithm

An apparatus for determining and correcting for background color of a document, comprising:

• a scanning module configured to obtain a digital representation of the document, the digital representation comprising a two-dimensional array of color elements, the color of each of which is specified by the combination of three color component values, the range of values of the color components defining a input color space containing a plurality of colors;
• and
• a processor configured to:
• quantize the color component values of the color elements, divide the input color space into a set of bins, each of which includes a subset of colors in the input color space, and assign each color in the digital representation to a particular one of the bins to obtain a summary of the distribution of colors in the digital representation,
• obtain, for each bin, statistical information that is indicative color component characteristics of colors in the bin,
• identify each bin that constitutes a color peak based on at least some of the statistical information,
• determine a region of influence for each color peak with respect to bins that are not color peaks,
• determine the color peak corresponding to the background color of the document according to predetermined decision-making rules that take into consideration the statistical information and region of influence determination, and
• construct a one-dimensional look-up table for each color component that correlates input color space values for that color component with corresponding output color space values.

A device-readable medium carrying a program of instructions for directing a machine to perform a method of determining and correcting for background color of a document, the program of instructions comprising:

• instructions for obtaining a digital representation of the document, the digital representation comprising a two-dimensional array of color elements, the color of each of which is specified by the combination of three color component values, the range of values of the color components defining a input color space containing a plurality of colors;
• instructions for (i) quantizing the color component values of the color elements, (ii) dividing the input color space into a set of bins, each of which includes a subset of colors in the input color space, and (iii) assigning each color in the digital representation to a particular one of the bins to obtain a summary of the distribution of colors in the digital representation;
• instructions for obtaining, for each bin, statistical information that is indicative color component characteristics of colors in the bin;
• instructions for identifying each bin that constitutes a color peak based on at least some of the statistical information;
• instructions for determining a region of influence for each color peak with respect to bins that are not color peaks;
• instructions for determining the color peak corresponding to the background color of the document according to predetermined decision-making rules that take into consideration the statistical information and region of influence determination;
• and
• instructions for constructing a one-dimensional look-up table for each color component that correlates input color space values for that color component with corresponding output color space values.