基于多尺度Retinex的图像增强算法

针对单景深低对比度图像中有用信息量少,可辨别性低的特点,提出运用多尺度Ret-inex理论对该类图像进行增强.根据单景深图像特点,在多尺度Retinex处理过程中采用平均标准偏差作为尺度参数的基础,设计多尺度参数高斯滤波器与原始图像在空域中进行滤波,最后运用灰度级均匀化对处理后图像进一步增强,结果显示处理效果良好.

Method of Image Enhancement Based on Multi-scale Retinex

The paper proposes image enhancement by multi-scale Retinex for signal depth of field which has the character of low contrast,little of available message and low of differentiability.According to the character of signal depth of field,multi-scale Retinex adopts multi-scale Gaussian filter filtering with original picture in space based on average standard deviation.At last,the paper deals with pictures by gray level equalization,the outcomes show that this method plays well.