Application of Improved Lzc Algorithm in the Discrimination of Photo and Text

Abstract

For the commonly used image two value method, there is a lack of precision and accuracy of the processing image containing noise or uneven illumination and so on, A new method for two valued difference image of complexity at different scales. Divide the signal sequence into several areas and do multi-scale binaryzation process to the sequence. Whith no increasing the number of symbols, get more meticulous and more accuracy complexity by the traditional LZC algorithm. The experiments show that the new method has advantages of certain anti-noise abilities, can be regarded as the feature of photo and text. Therefore, We can apply the improved LZC algorithm to the distinction between photo and text.