关于历史珍品残破图像快速修复仿真研究

Simulation of Fast Restoration of Broken Images of Historical Treasures

关于历史珍品残破图像的快速修复研究,在对历史珍品精确修复方面有重要意义。在图像的修复过程中,常因为磨损,腐蚀等外界因素,产生图像局部模糊。传统修复方法通过去噪处理降低图像模糊度,但未对局部纹理变化的阈值进行计算,导致修复精度较低。提出基于小波分析的历史珍品残破图像快速修复方法。该方法先融合于小波理论,将待修复的残破图像进行小波分解,计算出需要修复区域附近的纹理变化状态,得到待修复区域的纹理变化阈值,采用背景填充理论,初步修复待修复局部模糊区,对初步修复后的局部区域进行基于颜色匹配的快速修复,以此为依据实现历史珍品残破图像的快速修复。仿真证明,所提方法图像修复效果较好,符合人眼视觉要求,极大地提升了历史珍品残破图像的清晰度。

This paper presents a fast restoration method of broken image of historical treasure based on wavelet a- nalysis. Firstly, wavelet decomposition is used for the broken image to be repaired, then the state of texture change near the area to be repaired is calculated to get the threshold value of texture change in the area to be repaired. The background filling theory is used to repair the local fuzzy area preliminarily. Thus, the rapid restoration is carried out based on color matching for local area after preliminary, restoration. On this basis, the rapid restoration of broken im- age of historical treasures is realized. Simulation result proves that the proposed method has good effect of image res- toration and meets the human visual requirements, which greatly improves the clarity of broken image of historical treasures.