图像预处理中，通常采用直方图均衡化[1]和直

方图规定化、线性和非线性灰度变换法[2]提高图像

的对比度，但效果均不理想。故提出一种对比度增

强算法。该算法首先对直方图进行平滑处理以消除

直方图中因噪声而引入的随机干扰点，然后进行直

方图均衡化，增大图像的对比度，接着在整个显示

范围内对图像灰度级进行等间距排列，使输出图像

的灰度动态范围达到最大灰度变化范围，最后对整

幅图像中值滤波，滤除图像中被增强的噪声。-Image preprocessing, the commonly used histogram equalization [1] and histogram specification, linear and nonlinear gray level transformation method [2] to improve the image contrast, but the results are not satisfactory. It presents a contrast enhancement algorithm. In this algorithm the histogram is smoothed to eliminate the histogram due to the introduction of random noise and interference points, followed by histogram equalization, increases the contrast of the image, then in the context of the show were equally spaced gray level image arrangement, the output gray level dynamic range for maximum gray level range, and finally to the whole image median filter, filter out the noise image is enhanced.