基于全局和局部多特征的图像增强算法

Image enhancement algorithm based on global and local multi features

为了解决图像增强时存在的过度增强、图像信息量不足、性能退化等问题,本文提出了一种基于图像局部亮度平均值和全局对比度的图像增强算法。该算法根据图像局部亮度平均值来控制图像的局部对比度,并在全局部分采用局部与全局的对比度比例作为幂次变换的伽马值来进行图像增强。将图像分为低通滤波和高通滤波两部分,低通滤波分量控制高通滤波分量的幅度以增加局部对比度,然后对低通分量以非线性的方式修改图像的局部亮度平均值,并将其与处理后的高通分量进行融合。在对比实验中,本文提出的算法比次优算法的清晰度提高了9.9%,均方根误差(RMSE)提高了16%。实验证明本文算法在图像增强过程中能提升图像的清晰度,丰富图像细节,满足图像增强算法的稳定性好、鲁棒性强、清晰度高等要求。

In order to solve the problems of over enhancement and performance degradation in image enhancement,an image enhancement algorithm based on the local brightness average and global contrast of image is proposed.The algorithm controls the local contrast of image according to the local brightness average of image,and adopts the local and full contrast in global part.The local contrast ratio is the gamma value of the power transformation.Firstly,the image is divided into two parts:low pass filter and high pass filter.The low pass filter controls the amplitude of high pass filter to increase the local contrast.Then′the low-pass component is modified in a non-linear way and fused with the processed high-pass component.In the contrast experiment,the clarity of the algorithm proposed in this paper is 9.9%higher than that of the suboptimal algorithm,and the root mean square error(RMSE)is 16%higher.In the process of image enhancement,this algorithm can improve the definition of image,enrich the details of image,and meet the requirements of stability,robustness and high definition of image enhancement algorithm.