基于Richardson-Lucy算法的藏式壁画图像修复技术研究

Research on Image Restoration Technology of Tibetan Murals Based on Richardson-Lucy Algorithm

藏式壁画作为我国壁画文化中的重要组成部分,是我国民族艺术的瑰宝。随着时间的不断推移,由于受自然环境与保存条件的影响,壁画图像会出现不同程度的损坏。传统壁画图像修复技术的成本较高,且在图像转换与结构分解方面存在一定的缺陷。基于此,笔者提出基于Richardson-Lucy算法的藏式壁画图像修复技术。首先,全面分析藏式壁画图像结构与特点,获取图像损坏的相关信息;其次,采用逆波预处理方式生成去噪的壁画图像,构建藏式壁画图像模糊模型,估计图像先验的模糊核;最后,采用Richardson-Lucy算法修复破损的藏式壁画图像。实验表明,该藏式壁画图像修复技术执行的时间较短,且图像修复后的峰值信噪比(Peak Signal to Noise Ratio,PSNR)较高,视觉效果更具有优势。

As an important part of Chinese mural culture,Tibetan murals are treasures of Chinese national art.Over time,Tibetan murals have been affected by the natural environment and preservation conditions,and the mural images have been damaged to varying degrees.The traditional mural image restoration technology has high construction cost and has certain defects in image conversion and structural decomposition.Based on this,a research on Tibetan mural image restoration technology based on Richardson-Lucy algorithm is proposed.Firstly,,an all-round analysis of the image structure and characteristics of Tibetan murals was carried out to obtain relevant information about image damage;the inverse wave preprocessing was used to generate denoised mural images;Secondly,a Tibetan mural image blur model was constructed to estimate the image.Finally,the Richardson-Lucy algorithm is used to repair the damaged Tibetan mural image.Experiments show that this Tibetan mural image restoration technology takes a short time to perform,and the Peak Signal to Noise Ratio(PSNR) value after image restoration is higher,and the visual effect is more advantageous.