摘要
设计了一种基于机器视觉软件MVTec HALCON与图像拼接技术的文物修复系统.该系统首先对图像进行去噪,分割等预处理,然后利用Harris算法进行角点检测提取特征点,接着检测出的角点经归一化互相关匹配后使用RANCAC算法剔除误匹配点实现初步拼接,最后用改进的加权融合方法进行图像融合并利用图像修复技术进行修复.针对采用的算法编制了相应的软件,并进行了拼接实验,结果表明该方法可以快速有效地对图像进行无缝拼接,精度高,可广泛应用到古字画等文物修复工作中。
A conservation system designed based machine vision software MVTec HALCON and image stitching technology. In the system image is pretreated such as denoising, segmentation, and then use the Harris algorithm for comer detection feature extraction point, then detected comer points are normalized cross-correlation matching RANCAC algorithm to eliminate the mismatching points preliminary stitching, and finally with the weighted fusion method improved image fusion and image inpainting techniques to repair. Preparation for the algorithm used in the software and splicing experiment results show that the method can quickly and efficiently seamless, high accuracy, and can be widely applied to the restoration work of ancient calligraphy and other cuhural artifacts.
出处
《电子设计工程》
2013年第9期24-26,共3页
Electronic Design Engineering
基金
陕西省科技计划研究项目(12JK1004)
延安市科学技术研究发展计划项目(2010kg-09)
