Feature Extraction On Detecting Target In Clutter Background With Wavelet Transform

The wavelet transform has a lot of uses in the field of optical information processing. It's an effective technology applied to joint transform correlator system. It realizes the detection, identification and positioning of target. Joint transform correlator has not only high speed for identification but also high accuracy for positioning. Actually, it is difficult for joint transform correlator to identify target because the target image information which is collected by the imaging sensor has a lot of background noises. So we should depend on the technology of digital image processing, realize the correlation detection and identification of the target.In the processing of digital image, the edge represents the elementary character of image. Image extraction is one main resort on image management and analysis, and image extracted good or not that will affect the purpose of it. Formerly, there were a number of conventional methods, for instance, Robert,Sobel,Prewitt and Kirsh. They are all prone to operation, because of the complexity of image edge extraction, they could not get the balance between noise-proof feature and edge positioning. Because both the edge and noise are high frequency signal, it conduced that we make choice difficult between noise and edge.This paper has discussed the application of Gauss wavelet function on identifying target in clutter background. Wavelet transform is based on its multi-resolution character which can largely enhance the information of an image. On the basic of a mass of referenced literature, we know that Gauss wavelet function is familiar and effective on image extraction. Computer simulation results showed that the proposed method increases the intensity of the correlation point which can determine the object position accurately, especially under noisy environments. Optical experimental results also showed that the image feature extraction performance of the proposed algorithm is effective and competitive to other image processing algorithm reported in the literature.