A New Approach for Infrared Image Contrast Enhancement

摘要

In this paper, we propose a new algorithm to enhance global and local contrast of infrared image. The aim of our research is to carry out pretreatment for infrared video moving objects tracking. For the inherent difficulty, it is difficult for infrared image enhancement to get comparatively ideal result by adopting only one kind of method. So we propose a novel algorithm. First an enhancement algorithm is proposed based on plateau histogram using a self-adaptive threshold to enhance global contrast, which is different from the traditional histogram equalization algorithm. Second step, after getting a nonlinear gain function, the equalized infrared image is transformed by discrete stationary wavelet. Then the high frequency sub-bands are enhanced with the gain function better than linear filter. It also can suppress the amplification of noise. Experimental results show that the new algorithm can enhance the contrast of infrared image effectively and get more excellent visual effect than traditional histogram equalization method and unsharp masking method.