Research On The Automatic Classification Algorithm Of The Basic Elements In Chinese Landscape Paintings

Traditional Chinese painting (TCP) is precious cultural heritage of our country and is striding into the collection market of the world. But many imitations are badly affecting the internationalization of TCP. The traditional authentication method, which is strongly dependent on the experts' empirical knowledge, is usually lack of objective and convincing proofs. The computer-aided authentication of TCP based on recognition technology will help improve the reliability and scientificity. This thesis addresses itself to the authentication of TCP problem.In the first part of the article, we not only simply describe several aspects of TCP, such as the value, characteristics, classification and the trend of development, but also expatiate on three current authentication methods. Due to its shortcomings and to meet the current demand, We at last put forward a computer-assitance authentication method based on expert.In order to help experts compare and search details of the picture, we advance an automatic. classification method on the existing computer-assitance authentication platform. The new method can distinguish TCP's basic elements more credibly than before, the main jobs it involves are as follows:In the second chapter, we provide the classification of TCP's sub-objects, and describe the effective features according to the characteristics of Houses, Seals, Persons, Lichen-points and branches.In the third chapter, several classifiers are used to categorize the TCP's sub-objects in their feature space with 11 dimensionalities. To reach the best effect, Fisher and SVM are got to two levels of a decision tree.In the forth chapter, to make it more convenient, more reasonable and faster, we improve the automatic classification algorithm, and make further study on some aspects, such as TCP lookup, seal comparison and use modes.