Research on Real-time Images of Moving Targets Identification Based on the Fast Discrete Curvelet Transform

Abstract

This paper presents an algorithm of real-time image of moving targets identification based on the fast discrete Curvelet transform. Curvelet, as a new multiscale analysis algorithm, is more appropriate for the analysis of the image edges such as curve and line characteristics than wavelet, and it has better approximation precision and sparsity description. Introduced the curvelet transform to image processing, characteristics of original images are taken better and information for feature extraction is obtained more. The proposal of the second generation curvelet theory makes it to be understood and implemented more easily. Then the fast discrete curvelet transform based image moving targets identification method is proposed. Firstly, the source images are decomposed using curvelet transform, then extract the feature using the edge detection, and finally, get the results through the morphological analysis. Results of simulation experiment show the new method is speedy and very robust.