Abstract: Most of local histogram equalization algorithms used for contrast enhancement recently treat mean and median brightness as the dividing point of the image, these methods can keep the image brightness, but produce local excessive enhancement as well. So an adaptive local contrast enhancement algorithm with minimum brightness error is proposed, which preserves the brightness by choosing the best brightness split points based on absolute mean brightness error, then uses brightness preserving bi-histogram equalization algorithm equalizing sub-images of being split, and reduces blocking artifacts by filtering. The experimental results show that the algorithm not only preserves the brightness of the image, and achieves the local contrast enhancement as well.

Key words: brightness preservation, histogram equalization, adaptive contrast enhancement

摘要： 目前保持亮度的局部直方图均衡算法用于对比度增强时，大多以亮度均值和中值为图像的亮度分割点，这些方法能较好地保持图像的亮度，但同时也会产生局部过增强。为此提出了一种亮度误差最小的自适应局部对比度增强算法，根据亮度均值绝对误差自适应的选择最佳亮度分割点，然后用保持亮度的双直方图均衡算法对被分割的子图像进行均衡，最后用滤波器消除块效应。实验结果表明，该算法不仅保持了输入图像的亮度，同时也实现了局部对比度增强。

关键词: 亮度保持, 直方图均衡, 自适应对比度增强