基于对比度拉伸和梯度域优化的彩色图像灰度化算法

Decolorization Algorithms Based on Contrast Stretching and Gradient Domain Optimization

针对彩色图像灰度化无法较好保持原图像的局部对比度信息和结构的问题,提出了基于对比度拉伸和梯度域优化的彩色图像灰度化算法。首先,将图像分解为R、G、B共3个通道,并拉伸每个通道的对比度。其次,将拉伸后的图像合并进行初步灰度化。最后,将原图像的对比度信息与初步灰度化的结果图像结合构成能量函数,求得所需的灰度化图像。实验结果表明,该算法能较好地保持图像的结构和对比度信息。

For the problem that previous decolorization algorithms can not preserve local contrast information and structure,this paper propose decolorization algorithms based on contrast stretching and gradient domain optimization.First,decompose the image into R,G,and B channels,and stretch the contrast for each channel.Then combined the stretched R,G,and B channels’images and transform the stretched image from color to gray.At last,combine the contrast information of original image with the initial decolorization image to build the energy function,and obtain the result gray image.The experiment indicates that this algorithm can preserve the image structure and contrast information well.