一种增强图像对比度与局部细节的算法研究

Research on an algorithm for enhancing image contrast and local details

本文提出了一种简单而有效的图像增强方法来提高待处理图像的全局对比度和局部细节。首先,基于直方图均衡化框架,加入对数运算实现全局对比度增强。其次,采用拉普拉斯金字塔分解全局增强图像,加权每一层图像并叠加所有层,增强局部细节,并获得最终增强结果。实验表明,本文算法得到的图像主观视觉效果更自然,增强结果可比拟甚至优于目前最先进的基于一维和二维直方图的增强算法所得到的结果。

This paper proposes a simple and effective image enhancement algorithm to improve the global contrast and local details of the image to be processed.First,based on the histogram equalization framework,logarithmic operations are added to achieve global contrast enhancement.Secondly,the Laplacian pyramid is used to decompose the global enhanced image,weighting each layer of image and superimposing all layers to obtain the final enhancement results.Experiments show that the subjective visual effects of the images obtained by this algorithm are more natural,and the enhancement results can be compared to or even better than the results of the most advanced enhancement algorithm based on one-dimensional and two-dimensional histograms.