Abstract: The cause of amalgamation of gray level during the histogram equalization is analyzed through theory，the simple physical explain and analytical deduction are given，and the shortcoming of the algorithm is discussed；The algorithm of histogram equalization based on reserving gray level is brought up in order to reserve more detail information，eliminate the false outline，maintain the continuity and increase contrast of the image after processing，and the basic principle is analyzed；Through the result for experiment of PCB image that obtained by CCD and microscope verifies that the image after processing by the algorithm in this paper is better than that of the algorithm used by the histogram equalization.

摘要： 从理论上分析了直方图均衡化过程中灰度级合并的原因，给出了简单的物理解释和解析推导，并讨论了该算法的缺点；为了使图像保留更多的细节信息、消除虚假轮廓、保持连续性、提高对比度，提出了保留灰度级的直方图均衡化算法，并分析了其基本原理；对CCD与显微境获取的PCB图像进行了实验，结果表明：通过该算法处理的图像比采用传统直方图均衡化算法处理的图像的效果好。