Study Of Infrared Image Enhancement Methods Based On Spatial Domain

Infrared image enhancement is an important and basic technique in ART and Infrared Image Target Tracking System. Infrared image is characterized by low-contrast, blur and high-noise. These features will make a very bad impact on subsequent processing, for example, feature extraction, target recognition and tracking. Thus, it is very important to enhance infrared image.The author, in this paper, attempts to study infrared image enhancement of spatial domain through the perspective of studying infrared image contrast enhancement, noise reduction and infrared image edge enhancement. This paper will be focus on the follows:The producing principle and features of infrared image are studied. In order to reduce the amount of computation of histogram and satisfy the need of real-time system, a kind of fast method is proposed, i.e., interlace row/column histogram statistics.Common infrared image gray transform algorithms are studies in order to enhance infrared image contrast. The advantages and disadvantages of these approaches are compared by a large number of experiments. Based on the research, an algorithms, adaptive piecewise linear transform, is presented.After types, characters and models of noise are analyzed, the basic theory of image filtering enhancement on spatial domain is studied. On the basis of the above, common infrared image smoothing algorithms are studied and compared by experiment. Then, a kind of improved algorithms of gradient Inverse Weighting Smoothing and fast median filter algorithms are presented.Gradient algorithm and laplacian algorithm are analyzed and studied. Based on the experiment, two important sharpening algorithms are imposed. The experiment shows that the algorithms are effective. At last, the improvement laplacian is proved that it is effective.