An efficient method for automatic target position detection system

An automatic position detection system of a desired target is developed based on the nonlinear correlation approach. The hidden target is extracted by ortho-normalization and normalized correlation between camouflaged scene and target vectors. Uniform intensity transformation technique is applied for both of them. The camouflaged targets can be detected independently of the illumination conditions. Target positions can be detected not only for the same size of the target in a scene but also detect the various sizes and various oriented target included in scene. Furthermore, target positions from a camouflaged scene are detected in both noiseless and noisy environment. The effectiveness of the proposed system can be confirmed through the results of several experiments.