基于PageRank的SLAM后端优化研究
|
摘要
针对室内复杂非机构化环境建图及定位效率低的问题,提出了一种基于PageRank的同时定位与地图构建(SLAM)方法.在室内复杂非结构化环境中,SLAM前端建立的位姿图中包含大量待优化节点,根据SLAM后端的稀疏矩阵,利用PageRank算法对位姿图中节点进行筛选和排序,将低于设定阈值的节点在位姿图中剔除,保留与其他节点有高关联性的节点,减少位姿图中的节点,同时保留SLAM后端的稀疏特性,有效提高SLAM后端优化效率.在RGB-D标准数据集上进行实验验证,实验结果表明:在室内环境下,该SLAM后端优化算法缩短了优化时间,提高了实时性,且误差变化在可接受的范围内,为SLAM后端优化低效率问题提供了解决方案.
A fast method based on PageRank for the simultaneous localization and mapping(SLAM) was presented to address the problem of indoor complex and non-institutional environment mapping and low positioning efficiency.In the indoor complex unstructured environment, the pose graph created by the SLAM frontend contained a large number of nodes to be optimized.According to the sparse matrix created by the SLAM backend,the PageRank algorithm was used to filter and sort the nodes in the pose graph,and the nodes below the set threshold were removed from the graph,and the nodes with high correlation with other nodes were kept.The method reduced the nodes in the pose graph and retained the sparse characteristics of the SLAM backend,thus,SLAM backend optimization efficiency was improved.Experimental results were verified on the RGB-D standard dataset,which shows that the proposed SLAM backend optimization algorithm shortens the optimization time and improves the real-time performance in the indoor environment,and the error variation is within the acceptable range.The proposed method provides a solution to the problem of SLAM backend optimization in efficiency.
A fast method based on PageRank for the simultaneous localization and mapping(SLAM) was presented to address the problem of indoor complex and non-institutional environment mapping and low positioning efficiency.In the indoor complex unstructured environment, the pose graph created by the SLAM frontend contained a large number of nodes to be optimized.According to the sparse matrix created by the SLAM backend,the PageRank algorithm was used to filter and sort the nodes in the pose graph,and the nodes below the set threshold were removed from the graph,and the nodes with high correlation with other nodes were kept.The method reduced the nodes in the pose graph and retained the sparse characteristics of the SLAM backend,thus,SLAM backend optimization efficiency was improved.Experimental results were verified on the RGB-D standard dataset,which shows that the proposed SLAM backend optimization algorithm shortens the optimization time and improves the real-time performance in the indoor environment,and the error variation is within the acceptable range.The proposed method provides a solution to the problem of SLAM backend optimization in efficiency.
引文
[1]罗元,苏琴,张毅,等.基于优化RBPF的同时定位与地图构建[J].华中科技大学学报:自然科学版,2016,44(5):30-34.
[2]丛楚滢,王从庆,丁臻极,等.一种小型无人机的FastSLAM算法[J].华中科技大学学报:自然科学版,2015,43(S1):420-423,427.
[3]贾松敏,王可,郭兵,等.基于RGB-D相机的移动机器人三维SLAM[J].华中科技大学学报:自然科学版,2014,42(1):103-109.
[4]SUNDERHAUF N,PROTZEL P.Towards a robust backend for pose graph slam[C]//Proc of IEEE International Conference on Robotics and Automation.Saint Paul:IEEE,2012:1254-1261.
[5]JOHANNSSON H,KAESS M,FALLON M,et al.Temporally scalable visual slam using a reduced pose graph[C]//Proc of IEEE International Conference on Robotics and Automation.Karlsruhe:IEEE,2012:54-61.
[6]HE Z,YE C.An indoor wayfinding system based on geometric features aided graph slam for the visually impaired[J].IEEE Transactions on Neural Systems&Rehabilitation Engineering,2017,25(9):1592-1604.
[7]CADENA C,CARLONE L,CARRILLO H,et al.Past,present,and future of simultaneous localization and mapping:toward the robust-perception age[J].IEEETransactions on Robotics,2016,32(6):1309-1332.
[8]张洪华,刘璇,陈付豪,等.基于图优化的SLAM后端优化研究与发展[J].计算机应用研究,2019(1):1-9.
[9]DAVISON A J,REID I D,MOLTON N D,et al.Monoslam:real-time single camera slam[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2007,29(6):1052.
[10]THRUN S,MONTEMERLO M.The graph slam algorithm with applications to large-scale mapping of urban structures[J].International Journal of Robotics Research,2006,25(5):403-429.
[11]KAESS M,RANGANATHAN A,DELLAERT F.Isam:incremental smoothing and mapping[J].IEEE Transactions on Robotics,Roma,2008,24(6):1365-1378.
[12]KAESS M,JOHANNSSON H,ROBERTS R,et al.Isam2:incremental smoothing and mapping using the bayes tree[J].International Journal of Robotics Research,2012,31(2):216-235.
[13]XIE L,WANG S,MARKHAM A,et al.Graphtinker:outlier rejection and inlier injection for pose graph slam[C]//Proc of International Conference on Intelligent Robots and Systems.Vancouver:IEEE,2017:6777-6784.
[14]ROSEN D M,KAESS M,LEONARD J J.An incremental trust-region method for robust online sparse least-squares estimation[C]//Proc of IEEE International Conference on Robotics and Automation.Miami:IEEE,2014,30(5):1091-1108.
[15]CARLONE L,CENSI A,DELLAERT F.Selecting good measurements via?1,relaxation:a convex approach for robust estimation over graphs[C]//Proc of International Conference on Intelligent Robots and Systems.Miami:IEEE,2014:2667-2674.
[16]KAESS M,RANGANATHAN A,DELLAERT F.isam:incremental smoothing and mapping[J].IEEE Transactions on Robotics,2008,24(6):1365-1378.2008.
[17]KUMMERLE R,GRISETTI G,STRASDAT H,et al.G2o:a general framework for graph optimization[C]//Proc of IEEE International Conference on Robotics and Automation.Shanghai:IEEE,2011:3607-3613.
[18]任晓龙,吕林媛.网络重要节点排序方法综述[J].科学通报,2014,59(13):1175-1197.
[19]ENDRES F,HESS J,STURM J,et al.3D mapping with an RGB-D camera[J].IEEE Transactions on Robotics,Singapore,2017,30(1):177-187.
[2]丛楚滢,王从庆,丁臻极,等.一种小型无人机的FastSLAM算法[J].华中科技大学学报:自然科学版,2015,43(S1):420-423,427.
[3]贾松敏,王可,郭兵,等.基于RGB-D相机的移动机器人三维SLAM[J].华中科技大学学报:自然科学版,2014,42(1):103-109.
[4]SUNDERHAUF N,PROTZEL P.Towards a robust backend for pose graph slam[C]//Proc of IEEE International Conference on Robotics and Automation.Saint Paul:IEEE,2012:1254-1261.
[5]JOHANNSSON H,KAESS M,FALLON M,et al.Temporally scalable visual slam using a reduced pose graph[C]//Proc of IEEE International Conference on Robotics and Automation.Karlsruhe:IEEE,2012:54-61.
[6]HE Z,YE C.An indoor wayfinding system based on geometric features aided graph slam for the visually impaired[J].IEEE Transactions on Neural Systems&Rehabilitation Engineering,2017,25(9):1592-1604.
[7]CADENA C,CARLONE L,CARRILLO H,et al.Past,present,and future of simultaneous localization and mapping:toward the robust-perception age[J].IEEETransactions on Robotics,2016,32(6):1309-1332.
[8]张洪华,刘璇,陈付豪,等.基于图优化的SLAM后端优化研究与发展[J].计算机应用研究,2019(1):1-9.
[9]DAVISON A J,REID I D,MOLTON N D,et al.Monoslam:real-time single camera slam[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2007,29(6):1052.
[10]THRUN S,MONTEMERLO M.The graph slam algorithm with applications to large-scale mapping of urban structures[J].International Journal of Robotics Research,2006,25(5):403-429.
[11]KAESS M,RANGANATHAN A,DELLAERT F.Isam:incremental smoothing and mapping[J].IEEE Transactions on Robotics,Roma,2008,24(6):1365-1378.
[12]KAESS M,JOHANNSSON H,ROBERTS R,et al.Isam2:incremental smoothing and mapping using the bayes tree[J].International Journal of Robotics Research,2012,31(2):216-235.
[13]XIE L,WANG S,MARKHAM A,et al.Graphtinker:outlier rejection and inlier injection for pose graph slam[C]//Proc of International Conference on Intelligent Robots and Systems.Vancouver:IEEE,2017:6777-6784.
[14]ROSEN D M,KAESS M,LEONARD J J.An incremental trust-region method for robust online sparse least-squares estimation[C]//Proc of IEEE International Conference on Robotics and Automation.Miami:IEEE,2014,30(5):1091-1108.
[15]CARLONE L,CENSI A,DELLAERT F.Selecting good measurements via?1,relaxation:a convex approach for robust estimation over graphs[C]//Proc of International Conference on Intelligent Robots and Systems.Miami:IEEE,2014:2667-2674.
[16]KAESS M,RANGANATHAN A,DELLAERT F.isam:incremental smoothing and mapping[J].IEEE Transactions on Robotics,2008,24(6):1365-1378.2008.
[17]KUMMERLE R,GRISETTI G,STRASDAT H,et al.G2o:a general framework for graph optimization[C]//Proc of IEEE International Conference on Robotics and Automation.Shanghai:IEEE,2011:3607-3613.
[18]任晓龙,吕林媛.网络重要节点排序方法综述[J].科学通报,2014,59(13):1175-1197.
[19]ENDRES F,HESS J,STURM J,et al.3D mapping with an RGB-D camera[J].IEEE Transactions on Robotics,Singapore,2017,30(1):177-187.