An Algorithm of Feather and Down Target Detection and Tracking Method Based on Sparse Representation

Abstract

Feather and down products are becoming more and more popular. In recent years, the research team has done the down automatic detection and recognition by computer. In order to raise the recognition rate of feather and down category, in the paper a goal down image detection adaptive sparse representation and tracking method based on image is proposed, and target detection is used to determine the feather and down category. First of all, after the study, proposed improved adaptive sparse expression theory, and related features in the study of down on the image, the adaptive sparse representation theory and the multi-scale geometric analysis theory is applied to image recognition algorithm. The research will build a new adaptive sparse expression theory system, improve the robustness of the feather of image recognition, and will promote the target tracking technology innovation.