基于平均梯度的大幅壁画拼接图像清晰度评估研究  

Research on Clarity Evaluation of Large Scale Mural Stitching Images Based on Average Gradient

作　　者：安慧莉 王志明[3] 刁常宇[4] AN Huii;WANG Zhiming;DIAO Changyu(Dunhuang Academy,Dunhuang 736200;Cultural Heritage Digitization Institute,Dunhuang Academy,Mogao Grottoes,Dunhuang 736200;Nanjing University of Science and Technology,Nanjing 210094;Zhejiang University,Hangzhou 310058)

机构地区：[1]敦煌研究院,敦煌736200 [2]敦煌研究院国家古代壁画与土遗址保护工程技术研究中心,敦煌736200 [3]南京理工大学,南京210094 [4]浙江大学,杭州310058

出　　处：《计算机与数字工程》2024年第4期1173-1179,共7页Computer & Digital Engineering

基　　金：国家重点研发计划(编号:2020YFC1523002)资助。

摘　　要：图像拼接技术(Image Splicing)是数字图像处理的关键技术之一,在大幅壁画采用正射散点(Orthophoto Scatter-ing)的方法采集后,通过图像拼接技术形成大幅壁画数字图像,为确保拼接图像与原始拍摄图像的质量偏差在可控范围内,需对壁画图像清晰度进行评估。论文采用平均梯度(Average Gradient,AG)算子对图像清晰度进行计算分析,并设计了大幅壁画拼接图像清晰度评估流程;通过划分梯度的方式来表征局部图像清晰度层次,对壁画图像清晰度进行递增和递减运算,得出壁画原始拍摄图像清晰度真值,通过对不同洞窟体积、建筑结构、壁画面积和壁面曲率壁画原始图像平均梯度值计算,确定了壁画原始图像平均梯度值的区间;提出了拼接融合图像平均梯度值的获取方法。以莫高窟第172窟大幅壁画数字图像为例,开展评价函数计算实验,分析比对原始图像平均梯度值与拼接融合图像清晰度值的差异,实验结果表明:平均梯度算子作为客观评价指标,在壁画拼接结果的清晰度评估方面具有可行性。Image stitching technology is one of the key technologies in digital image processing.After collecting large murals using the Orthophoto Scattering methods,image stitching technology is used to form digital images of large murals.The clarity of the mural images needs to be evaluated to ensure that the quality deviation between the stitched images and the original captured images is controllable.In this paper,the Average Gradient(AG)operator is used to calculate and analyze the image sharpness,and the process of image sharpness evaluation is designed,the clarity of murals is calculated by increasing and decreasing,and the true val-ue of the original murals is obtained,by calculating the average gradient values of the original murals with different cave volumes,architectural structures,murals area and wall curvature,the interval of the average gradient values of the original murals is deter-mined.The method of obtaining the average gradient value of Mogao Grottoes mosaic fusion image is put forward,and the evaluation function of the large-scale murals in Cave 172 of Mogao Grottoes is calculated,the results show that it is feasible to use the average gradient operator as an objective evaluation index to evaluate the clarity of fresco mosaic.

关 键 词：莫高窟壁画 数字化技术 图像拼接技术 清晰度 平均梯度 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]