基于视觉色彩改进的数字图像智能编修方法设计  

Design of intelligent restoration method for digital images based on visual color

作　　者：钱江 丁懿 QIAN Jiang;DING Yi(College of Art,SUST,Suzhou 215011,China;School of Art&Design,Shaanxi University of Science&Technology,Xi'an 710021,China)

机构地区：[1]苏州科技大学艺术学院,江苏苏州215011 [2]陕西科技大学设计与艺术学院,陕西西安710021

出　　处：《苏州科技大学学报（自然科学版）》2024年第3期67-74,84,共9页Journal of Suzhou University of Science and Technology（Natural Science Edition）

基　　金：国家自然科学基金项目(61672371,61803279)。

摘　　要：为了实现更加精准、高效的图像修复,提高图像的质量和完整度,论文提出了基于视觉色彩的数字图像智能修复方法设计。首先,采用基于分通道自适应理论(ATSC)阈值分割方法将图像中的待修复区域准确、快速地分割出来,为后续的修复工作提供目标修复区域;其次,使用基于改进样本块的修复模型实现待修复区域的智能填补修复;最后,引入基于聚类的颜色迁移方法提高修复后的数字图像质量,使修复后的图像在色彩上更加丰富、饱满,提高图像色彩的整体视觉效果。实验结果表明,该方法具有较高的修复精度,能够有效增强修复后图像的整体质量和视觉效果。In order to achieve more accurate and efficient image restoration and improve image quality and completeness,we put forward a digital image intelligent restoration method design based on visual color.Firstly,the threshold segmentation method based on the adaptive theory of sub channels(ATsc)was adopted to accurately and quickly segment the area to be repaired in the image,providing a target repair area for subsequent work.Secondly,a repair model based on improved sample blocks was used to achieve intelligent filling and repair of the area to be repaired.Finally,a cluster-based color transfer method was adopted to improve the quality of the repaired digital image,making the repaired image rich and full in color,and improving the overall visual effect of image color.The experimental results show that the proposed method has high restoration accuracy and can effectively enhance the overall quality and visual effect of the repaired image.

关 键 词：ATSC阈值分割 图像修复 样本块修复 颜色迁移 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]