Lossless compression of color image data using entropy encoding

A method of decompressing compressed image data, the compressed image data including a data block including plural data bits and a code block including plural codes for each image line of pixels, each pixel having a color value, the method comprising the steps of:

• reconstructing the color values of the pixels by decoding successive codes in the code block, while
• for a current code,
• checking a code type denominator included in the current code for being one of a predetermined set of compression codes, and, if said code type denominator denominates the code as a compression code from the predetermined set,
• deducting a number from the code type denominator of the current code and reading that number of data bits from the data block,
• calculating a predicted color value, using a predictor, for a current pixel related to the current code, and
• reconstructing the color value of the current pixel from the predicted color value in combination with a difference value based on the data bits read from the data block.

A non-transitory computer-readable medium containing a computer program comprising computer-executable instructions for executing a method of compressing a digital continuous tone image of pixels situated on scan lines, each pixel having a pixel value defined by at least one color channel value, a color channel value having B bits of precision (B>1), the method producing compressed image data and, for a current pixel to be encoded, said current pixel having an actual pixel value, said method comprising the steps of:

• predicting a predicted pixel value based on pixel values of at least one previously processed pixel from the same image, using a fixed rule;
• determining a difference parameter based on a difference value of said predicted pixel value and the actual pixel value of said current pixel to be encoded;
• inspecting the difference parameter for existence of an uninterrupted series of highest order bits having a value equal to zero;
• and
• removing at least part of said highest order zero bits, and, if a number of bits within predetermined limits remains, generating a compression code having a predetermined fixed length, said code being selected from a predetermined set of compression codes based on the number of remaining non-zero bits.

An apparatus for compressing a digital continuous tone image of pixels situated on scan lines, each pixel having a pixel value defined by at least one color channel value, a color channel value having B bits of precision (B>1), the apparatus producing compressed image data and comprising:

• an encoding processor that encodes a current pixel, said current pixel having an actual pixel value;
• a predictor processor that predicts a predicted pixel value based on pixel values of at least one previously processed pixel from the same image, using a fixed rule;
• a difference parameter determinator processor that determines a difference parameter based on a difference value of said predicted pixel value and the actual pixel value of said current pixel to be encoded;
• a checking processor that inspects the difference parameter for existence of an uninterrupted series of highest order bits having a value equal to zero;
• and
• a cutting-off processor that removes at least part of said highest order zero bits, and, if a number of bits within predetermined limits remains,
• a compression code generator processor that generates a compression code having a predetermined fixed length, said code being selected from a predetermined set of compression codes based on the number of remaining non-zero bits.