基于凸优化改进的相机全局位置估计方法  

An Improved Method for Camera Location Estimation Through Convex Optimization

作　　者：谢理想[1] 万刚[1] 曹雪峰[1] 王庆贺[1] 王龙[2] XIE Li-Xiang1 ,WAN Gang1, CAO Xue-Feng1 ,WANG Qing-He1 ,WANG Long2(1. PLA Information Engineering University, Zhengzhou 45000 2. National Defense Information College, Wuhan 43000)

机构地区：[1]解放军信息工程大学,郑州450000 [2]国防信息学院,武汉430000

出　　处：《自动化学报》2018年第3期506-516,共11页Acta Automatica Sinica

基　　金：国家自然科学基金(41301428;4110143);地理信息工程国家重点实验室开放基金(SKLGIE2016-M-3-4)资助~~

摘　　要：相机全局位置估计作为运动恢复结构算法(Structure from motion,SfM)中的核心内容一直以来都是计算机视觉领域的研究热点.现有相机全局位置估计方法大多对外点敏感,在处理大规模、无序图像集时表现的尤为明显.增量式SfM中的迭代优化步骤可以剔除大部分的误匹配从而降低外点对估计结果的影响,而全局式SfM中没有有效地剔除误匹配的策略,估计结果受外点影响较大.针对这种情况,本文提出一种改进的相机全局位置估计方法:首先,结合极线约束提出一种新的对误匹配鲁棒的相对平移方向估计算法,减少相对平移方向估计结果中存在的外点;然后,引入平行刚体理论提出一种新的预处理方法将相机全局位置估计转化为一个适定性问题;最后,在此基础上构造了一个对外点鲁棒的凸优化线性估计模型,对模型解算获取相机位置估计全局最优解,本文方法可以很好地融合到当下的全局式SfM流程中,与现有典型方法的对照实验结果表明:在处理大规模、无序图像时,本文方法能显著提高相机全局位置估计的鲁棒性,并保证估计过程的高效性和估计结果的普遍精度.As a core module of structure from motion （SfM）, location estimation of cameras in a global framework has been a research hotspot of computer version. State-of-the-art methods for location estimation are sensitive to outliers, especially for large scale, unordered images. The incremental SfM reduces the influence of outliers through an iterative optimization. The global SfM does not have an efficient strategy to remove mismatch, so the result of estimation is influenced deeply by outliers. Therefore, we introduce an improved method for location estimation. First, combined with the epipolar constraint we propose a new pairwise direction estimation algorithm. Then, we make the problem well-posed by introducing a new preprocessing method based on parallel rigidity. Finally, we propose a robust linear estimation model based on convex programing. We can get a global optimum solution by resolving this model. The method can integrate well with state-of-art global SfM pipeline. Multiple group experiments have proved the robustness of our methods without any loss of efficiency and common precision.

关 键 词：日运动恢复结构 全局位置估计 平行刚体 凸优化 极线几何 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]