Database Classification of MPEG7 Shape Using Rotation Moments and Generated Stationary Transformed Features

After the main trellis processing, the final score will contain all matching information. The graph matching shape classification is also used, where the Extracted Stationary Transformed Features is used to describe the Gabor graph for signal Rotation Invariant Moments object. The difference between our work and theirs is that we use feature point’s as graph node values and they use Shape Database Classification Gabor jet. Another difference is that and they didn’t use to do the graph matching.