The layout recognition and reconstruction of Chinese documents based on Gabor filter

The layout recognition of Chinese documents is one of the most difficult pattern recognition problems, because it concerns a great number of classes and many variations. In this paper, an intelligent layout recognition and reconstruction system is put forward on the basis of the analysis of the present methods. The algorithms of the pivotal processing steps such as layout analysis and character font recognition are discussed. Especially, a 2D Gabor filter is used in character font recognition to improve the adaptability to the Chinese document. Experiments have been made and some promising results have been drawn.