局部特征与全局信息联合的自适应图像增强算法

针对传统基于修正直方图的图像增强算法不能兼顾局部特征和全局信息的问题,提出一种局部特征与全局信息联合的自适应图像增强算法. 该算法将增强分为局部增强和全局增强两部分,局部增强利用像素的邻域信息和局部与全局对比度的比例信息作为幂次变换的伽马值,对图像进行伽马校正,提高图像的亮度和局部对比度;全局增强利用区域相似直方图统计抑制噪声,避免过度增强. 实验结果表明,本文算法在客观性能上优于其它传统图像增强算法,并且可以有效提高复杂光照下人脸图像的检测率. 

Local-to-Global Adaptive Image Enhancement Algorithm

In traditional histogram modification technologies, there is a problem that the algorithm can't deal with the local features and global information simultaneously. In this paper, we proposed a new adaptive image enhancement algorithm which used both the local features and global information. Image enhancement was divided into local enhancement and global enhancement. In the step of local enhancement, the pixel neighborhood information and the ratio between local contrast and global contrast were used as the gamma value of a power transformation. This process increased the local contrast and the luminance of the dark region. Then, the region similarity histogram was developed in global enhancement to suppress the noise and avoid the over-enhancement. The experiments show that the proposed algorithm is better than traditional image enhancement methods, and it can improve the face detection ratio under complex illumination.