Eliminating of Image Matching Gross Errors Based on Local Vector Field
Eliminating of Image Matching Gross Errors Based on Local Vector Field
复制标题
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Zhang Yongju
中科院分区：
文献类型：
--
作者：
Zhang Yongju
本文提出了一种消除图像匹配中粗误差的方法。整个过程可分为三个步骤。首先，消除了制度差异。然后构造图像匹配点的不规则三角网（TIN），得到分割局部域；基于图像匹配粗误差的正态分布，提出了一种局部域统计向量描述符。最后，采用合理的阈值来消除粗差。通过两组数据的实验验证了所提方法的可行性。结果表明，处理速度快，总误差消除成功率高。该方法也为摄影误差处理和可靠性理论的研究提供了新的视角。
This paper proposes a method that can be applied to eliminating gross errors in image matching.The whole process can be divided into three steps.Firstly,the systematic difference is removed.Then triangulated irregular network(TIN)of image matching points is constructed to get the partitioning local field.Based on the normal distribution of the image matching gross error,a vector descriptor is proposed in the statistics on the local field.Finally,a reasonable threshold is used in eliminating gross errors.The feasibility of the proposed method is verified based on the experiments of two groups of data.The results showed high processing speed and success rate of gross error elimination.This method also provided a new viewpoint to the research of photographic error processing and reliability theory.