现有的直方图均衡算法在增强图像对比度的同时， 输出图像的亮度与输入图像无关， 并且在均衡区域产生亮度饱和现象，提出了一种新的直方图均衡算法。以亮度保持的双直方图均衡算法( B B HE )为基础，改进其对子图像的分类方式：根据直方图对图像进行多峰值分解，得到一系列不同范围的子图像，然后对每一个子图像在其相应的灰度范围内进行直方图均 衡， 最后合并这些子图像的均衡结果。实验结果表明，直方图均衡新算法不仅在保持了输出图像亮度的同时， 而且非常有效 的解决了在原图像均衡区域产生的亮度饱和问题对图像的影响。-The existing histogram balanced algorithm during enhancement image contrast gradient s, outgoing picture s brightness has nothing to do with the input picture, and has the brightness saturated phenomenon in the balanced region, proposed one kind of new histogram balanced algorithm. The double histogram balanced algorithm which maintains by brightness (B B HE) is a foundation, improves its antithetical couplet image the classified way: Carries on the multi-peak values according to the histogram to the image to decompose, obtains a series of different scope sub-image, then carries on the histogram to each sub-image in its corresponding density latitude to be balanced, finally merges these sub-image balanced result. The experimental result indicated that the histogram balanced new algorithm not only while maintained outgoing picture brightness, moreover the very effective solution the brightness saturated question which produced in the original image balanced region to the image the influ