基于Retinex理论的草地图像增强算法研究

Abstract

为了解决室外环境中由于光照不均或者大雾天气下，草地图像模糊不清，不能有效的识别和提取草地图像重要信息的问题，本文提出一种专门针对草地图像的图像增强算法。本文基于传统的Retinex理论，用小波变换的方法拆分图像的高频分量和低频分量分别进行处理，之后再将处理结果融合重构。实验结果表明，相比传统的MSRCR算法和直方图均衡化方法，本文的算法提高了图像清晰度并且有效的抑制了图像处理时产生的噪声。 The article proposes a kind of image enhancement algorithm for lawn image, in order to solve the problem that lawn image is so blurred due to uneven illumination or heavy fog weather that it can’t identify and extract the important information effectively in outdoor environments. This paper uses wavelet transform method based on the traditional Retinex theory. It split the high frequency component and low frequency component then deal with them separately. Finally, the results are fused and reconstructed. Experimental results show that the algorithm of this paper provides the image’s definition and controls noise in image processing when it compared with traditional MSRCR algorithm and histogram equalization algorithm.