Research And Implementation Of Color-to-gray Algorithm

The essence of color image graying is a process of reducing the dimension of a threedimensional pixel matrix of a color image to generate a two-dimensional gray matrix.During the conversion process,the salient feature information of the color image should be preserved to the maximum extent,so that the brightness,contrast and structure information of the image before and after conversion can be matched optimally.At present,the commonly used grayscale algorithms based on optimization of target energy function and image information extraction are easy to loss of contrast information.In view of the above defects,this paper explores and studies the local contrast enhancement color grayscale algorithm based on Contourlet transform and the linear projection grayscale algorithm based on structural information similarity.Aiming at the problem that the output gray image cannot accurately preserve the contrast information of the original image in the grayscale algorithm based on the optimized target energy function,this paper studies a local image contrast gray image graying algorithm based on Contourlet transform.The algorithm firstly performs initial grayscale on the input color image in CIE Lab color model.Then,the Contourlet transform is used to locally enhance the initial grayscale image to obtain the final grayscale image.The experimental results show that the proposed algorithm can effectively enhance the local contrast of grayscale images,maintain more contrast information of the original image,the subjective and objective evaluation results are better than some typical grayscale algorithms.Aiming at the loss of contrast information caused by discrete projection weights in linear projection grayscale method based on image information extraction,this paper studies a linear projection gradation algorithm based on structural similarity.Firstly,a contrast image is constructed in RGB color space according to the input image;secondly,the weight coefficients of the corresponding channels are calculated by using the structural similarity between the three single-channel images of the input image and the contrast image;finally,the grayscale image is obtained by using the weighted average.The experimental results show that the algorithm can effectively retain the contrast information of the original image and improve the computational efficiency,the subjective and objective evaluation results are optimal.