Heritage Image Identification Algorithm Research

In this thesis, the images of cultural relics are taken as research objects.On the base of consultation with the relevant experts and application ofpattern recognition and digital image processing technologies, throughincessant testing and optimization, we expound the technique based onshape and texture identification, and take multiple steps to identify theculture relics to accustom various requirements.Technique based on shape identification is using geometry constants,fourier descriptor, wavelet descriptor and so on. Technique based ontexture identification is using traditional Coocurrence feature andfractal dimensional feature, and feature based on wavelet analysis.From the experiments, we can see that the identification algorithm canmeet the basic requirements of cultural relic identification, and with thefurther development, it may be used in practice.