大视场红外图像的直方图增强算法研究

摘要

探索了大视场红外图像增强的研究方法,考虑到大视场下背景的差异性,将图像切割成几个子图像分别进行增强.同时,为了保持相邻子图像间的空间相关性,切割图像的时候保留一定的重叠区.然后对每个子图像采用自适应平台直方图增强算法处理,最后对生成的每个子图像增强结果进行拼接,重叠区部分采用图像加权融合方法进行处理.最后通过实验证明了提出方法的优势和可行性.%In this paper,we explore the wide-view infrared image enhancement algorithm. Considering the difference of background in wide view image, we divide the image into several sub-images which are enhanced respectively. At the same time,in order to maintain the space relativity between the adjacent sub-images,the image overlaps are retained. Then, each sub-image is processed using adaptive platform histogram enhancement. The sub-images enhanced are merged finally,and the image overlaps are fused by image weighted fusion method. At last we demonstrate the advantages and feasibility of the algorithm proposed in the paper by experiments.