Chinese handwriting-based writer identification by texture analysis

Handwriting-based writer identification is a hot research field in pattern recognition. Given that the handwritings of different people are often visually distinctive, we regard the global text and single character of Chinese as textures and thus apply techniques of texture analysis to them. We present two methods of texture analysis on Chinese off-line handwriting writer identification, one is text-independent and the other is text-dependent. Further, the combination of these two methods is also applied on our application. Through experiments, ideal results are achieved.