REUSE METHODE FOR DESCRIPTOR OF IMAGE

摘要

PURPOSE: A video descriptor reusing method is provided to reduce the operation of STIF algorithm by reusing a descriptor under specific conditions about parts of frames including backgrounds. CONSTITUTION: Key points are detected from a current frame by SIFT(Scale-Invariant Feature Transform). A descriptor for each key point is generated. The key points are generated in the next frame by the SITF. The key points are compared with each other. When key points generated in the next frame satisfy set reuse conditions, the descriptor generated in the current frame is reused in the next frame.