Color-to-gray Conversion Algorithm Study Based On Adaptive Linear Weighting

With the development of computer technology,digital media based on image is deepening and enriching our life.Color image plays an important role in the process of transmitting information because of its rich color and good visual effect.The gray image can represent most of the features of the color image with less data information,and plays an important role in the fields of pattern recognition,image segmentation and printing.Color-to-gray conversion is a mapping process from three-dimensional space to one-dimensional space.In this process,the loss or weakening of characteristic information such as chromaticity,contrast and structure is inevitable.Therefore,how to keep the key information of the original image as much as possible has become the main research purpose.In recent years,researchers at home and abroad have proposed many methods to convert color image into gray image,among which linear weighting method is popular and simple.Because of its simple principle and good use of the whole gray level,it is still widely used.Based on this kind of method,the main contributions of this paper are as follows.This paper presents a new color-to-gray conversion algorithm based on adaptive linear weighting.Firstly,we discrete the weight space by exhaustive method and determine the candidate gray images.Then we calculate the number of each candidate gray image boundary points,which adaptively selects the optimal projection direction.The algorithm proposed in this paper focuses on the edge information of the image,and extracts the edge points of the image by Canny edge detection.Subjective evaluation and objective evaluation are carried out between this algorithm and the experimental results of other existing algorithms.In subjective evaluation,the conversion effect of this algorithm accords with the visual characteristics of human eyes in terms of the images with obvious regional features.C2G_SSIM and E-Score models are used in objective evaluation.The average score of the algorithm in thispaper is higher than that of the comparative algorithm in both models.For the case where the advantages and disadvantages of the algorithm cannot be distinguished in the objective analysis,we calculated the running time of the algorithm.The results show that the running time of the algorithm in this paper is shorter.