Abstract: Equalization is equal to uniform distribution, image enhancement algorithms directed by the equalization ideology have rich connotation, profound significance, wide application. Histogram Equalization（HE） is representative strongly in the field of image enhancement, on the basis, in recent years, researchers have expanded the applications of equalization ideology in the field of image enhancement. However, there is lack of the reviewing paper about HE for image enhancement which can comprehensively summarize HE related techniques, this paper considers the equalization ideology as the mainline, the improved work is classified into 5 kinds of techniques, meanwhile, the typical algorithms of each technique are analyzed, finally, the probable research directions in future are pointed out.

Key words: image enhancement, histogram equalization, sub-histogram equalization, histogram modification, variation technique, local histogram equalization, gradient domain, wavelet domain

摘要： 均衡化即等可能分布，以均衡化指导思想的图像增强方法内涵丰富，意义深远，应用广泛。直方图均衡化（Histogram Equalization，HE）算法在图像增强中具有很强的代表性，近几年研究人员在此基础上又进一步拓展了均衡化思想在图像增强领域的应用。不过，目前缺乏针对直方图均衡化技术较为全面的综述评论性文章，以均衡化思想为主线，对于相关改进工作，分为五类技术，并分析了各类技术的典型算法，最后指出了今后可能的研究方向。

关键词: 图像增强, 直方图均衡化, 子直方图均衡化, 直方图修正, 变分技术, 局部直方图均衡化, 梯度域, 小波域