高精度移动目标位姿测量方法
|
摘要
研究了微小型机器人运动中高精度位姿测量方法。针对核相关滤波(KCF)算法对快速移动目标跟踪中由于边界效应导致误差跳动与跟踪丢失的问题,提出了基于核相关滤波的自适应(AKCF)跟踪算法。该算法融合卡尔曼滤波器作为目标的位置预测器,通过核相关滤波检测实现目标加速度方差和频率自适应,完成对卡尔曼位置预测器的校准,建立预测-检测-校准的跟踪机制;对跟踪目标形态学处理、最小二乘法圆度拟合高精度的提取特征点;利用单目视觉标定原理构建运动坐标系与图像坐标系之间的映射关系,完成高精度位姿测量。跟踪算法仿真测试实验中,AKCF算法能够适应不同运动状态下目标的跟踪,解决KCF算法在跟踪过程中容易出现的目标漂移甚至丢失的问题;建立实验测试平台完成测量系统精度验证,在400 mm×300 mm视场内,对半径约为3 mm、帧间加速位移不超过5.6 mm的移动目标的位置检测的均方根误差达到0.0856 mm,姿态角度检测的均方根误差达到0.1246°。
This study focuses on the method of high-precision pose measurement for moving micro-robot.In order to solve the problem of error runout and tracking loss caused by boundary effect of kernelized correlation filter(KCF) algorithm when tracking the fast moving target, an adaptive kernelized correlation filter(AKCF) algorithm is proposed. To establish a prediction-test-calibration tracking mechanism, the Kalman filter is combined as the target position predictor and through the kernel correlation filter detection, acceleration variance and frequency of target is adapted to correct the position predictor. By processing the morphology of the tracking target, the least square method is used to fit the high precision feature points. Based on the principle of monocular vision calibration, the mapping relationship between the moving coordinate system and the image coordinate system is constructed to complete high precision pose measurement.The simulation results validate that AKCF algorithm can track the target in different states and solve the problem that the KCF algorithm is prone to drift or even lose the target in the tracking process. An experimental test platform is established to verify the accuracy of the measurement system. In the 400 mm×300 mm field of view, for the radius of target about 3 mm and the acceleration displacement between frames not exceeding the 5.6 mm, the root-mean-square error of position is 0.0856 mm, and the root-mean-square error of attitude angle is 0.1246 °.
This study focuses on the method of high-precision pose measurement for moving micro-robot.In order to solve the problem of error runout and tracking loss caused by boundary effect of kernelized correlation filter(KCF) algorithm when tracking the fast moving target, an adaptive kernelized correlation filter(AKCF) algorithm is proposed. To establish a prediction-test-calibration tracking mechanism, the Kalman filter is combined as the target position predictor and through the kernel correlation filter detection, acceleration variance and frequency of target is adapted to correct the position predictor. By processing the morphology of the tracking target, the least square method is used to fit the high precision feature points. Based on the principle of monocular vision calibration, the mapping relationship between the moving coordinate system and the image coordinate system is constructed to complete high precision pose measurement.The simulation results validate that AKCF algorithm can track the target in different states and solve the problem that the KCF algorithm is prone to drift or even lose the target in the tracking process. An experimental test platform is established to verify the accuracy of the measurement system. In the 400 mm×300 mm field of view, for the radius of target about 3 mm and the acceleration displacement between frames not exceeding the 5.6 mm, the root-mean-square error of position is 0.0856 mm, and the root-mean-square error of attitude angle is 0.1246 °.
引文
[1]Kelly I,Martinoli A.A scalable,on-board localisation and communication system for indoor multi-robot experiments[J].Sensor Review,2004,24(2):167-180
[2]Pugh J,Martinoli A.Relative localization and communication module for small-scale multi-robot systems[C].In:Proceedings of the 2006 IEEE International Conference on Robotics and Automation,Orlando,USA,2006.188-193
[3]Qazizada M E,Pivar8iováE.Mobile robot controlling possibilities of inertial navigation system[J].Procedia Engineering,2016,149:404-413
[4]张永顺,郭锐,刘煜,等.管内游动微型机器人的在线定位方法.哈尔滨工业大学学报[J],2004,36(10):1382-1384
[5]Chu H K,Mills J K,Cleghorn W L.Dynamic tracking of moving objects in microassembly through visual servoing[C].In:Proceedings of the 2010 IEEE International Conference on Mechatronics and Automation,Xi’an,China,2010.1738
[6]Chu H,Xie Z,Nie X,et al.Particle filter target tracking method optimized by improved mean shift[C].In:Proceedings of the 2010 IEEE International Conference on Information and Automation,Yinchuan,China,2014.991-994
[7]Du K,Ju Y,Jin Y,et al.MeanShift tracking algorithm with adaptive block color histogram[C].In:Proceedings of the International Conference on Consumer Electronics,Communications and Networks,Yinchuan,China,2012.2692-2695
[8]Dowson N,Bowden R.Mutual information for Lucas-kanade tracking(MILK):an inverse compositional formulation[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2007,30(1):180-185
[9]Kalal Z,Mikolajczyk K,Matas J.Face-TLD:trackinglearning-Detection applied to faces[C].In:Proceedings of the 2010 IEEE International Conference on Image Processing,Hong Kong,China,2010.3789-3792
[10]Bolme D S,Beveridge J R,Draper B A,et al.Visual object tracking using adaptive correlation filters[C].In:Proceedings of the Computer Vision and Pattern Recognition,San Francisco,USA,2010.2544-2550
[11]Henriques J F,Rui C,Martins P,et al.Exploiting the circulant structure of tracking-by-detection with kernels[C].In:Proceedings of the European Conference on Computer Vision,Florence,Italy,2012.702-715
[12]Dalal N,Triggs B.Histograms of oriented gradients for human detection[C].In:Proceedings of the 2005 IEEEComputer Society Conference on Computer Vision and Pattern Recognition,San Diego,USA,2005.886-893
[13]Henriques J F,Rui C,Martins P,et al.High-speed tracking with kernelized correlation filters[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2015,37(3):583-596
[14]Wu Y,Lim J,Yang M H.Object Tracking Benchmark[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2015,37(9):1834-1848
[15]崔彦凯,梁晓庚,贾晓洪,等.改进的机动目标当前统计模型自适应跟踪算法[J].计算机仿真,2013,30(3):42-44
[16]巴宏欣,何心怡,方正,等.机动目标跟踪的一种新的方差自适应滤波算法[J].武汉理工大学学报(交通科学与工程版),2011,35(3):448-452
[2]Pugh J,Martinoli A.Relative localization and communication module for small-scale multi-robot systems[C].In:Proceedings of the 2006 IEEE International Conference on Robotics and Automation,Orlando,USA,2006.188-193
[3]Qazizada M E,Pivar8iováE.Mobile robot controlling possibilities of inertial navigation system[J].Procedia Engineering,2016,149:404-413
[4]张永顺,郭锐,刘煜,等.管内游动微型机器人的在线定位方法.哈尔滨工业大学学报[J],2004,36(10):1382-1384
[5]Chu H K,Mills J K,Cleghorn W L.Dynamic tracking of moving objects in microassembly through visual servoing[C].In:Proceedings of the 2010 IEEE International Conference on Mechatronics and Automation,Xi’an,China,2010.1738
[6]Chu H,Xie Z,Nie X,et al.Particle filter target tracking method optimized by improved mean shift[C].In:Proceedings of the 2010 IEEE International Conference on Information and Automation,Yinchuan,China,2014.991-994
[7]Du K,Ju Y,Jin Y,et al.MeanShift tracking algorithm with adaptive block color histogram[C].In:Proceedings of the International Conference on Consumer Electronics,Communications and Networks,Yinchuan,China,2012.2692-2695
[8]Dowson N,Bowden R.Mutual information for Lucas-kanade tracking(MILK):an inverse compositional formulation[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2007,30(1):180-185
[9]Kalal Z,Mikolajczyk K,Matas J.Face-TLD:trackinglearning-Detection applied to faces[C].In:Proceedings of the 2010 IEEE International Conference on Image Processing,Hong Kong,China,2010.3789-3792
[10]Bolme D S,Beveridge J R,Draper B A,et al.Visual object tracking using adaptive correlation filters[C].In:Proceedings of the Computer Vision and Pattern Recognition,San Francisco,USA,2010.2544-2550
[11]Henriques J F,Rui C,Martins P,et al.Exploiting the circulant structure of tracking-by-detection with kernels[C].In:Proceedings of the European Conference on Computer Vision,Florence,Italy,2012.702-715
[12]Dalal N,Triggs B.Histograms of oriented gradients for human detection[C].In:Proceedings of the 2005 IEEEComputer Society Conference on Computer Vision and Pattern Recognition,San Diego,USA,2005.886-893
[13]Henriques J F,Rui C,Martins P,et al.High-speed tracking with kernelized correlation filters[J].IEEETransactions on Pattern Analysis&Machine Intelligence,2015,37(3):583-596
[14]Wu Y,Lim J,Yang M H.Object Tracking Benchmark[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2015,37(9):1834-1848
[15]崔彦凯,梁晓庚,贾晓洪,等.改进的机动目标当前统计模型自适应跟踪算法[J].计算机仿真,2013,30(3):42-44
[16]巴宏欣,何心怡,方正,等.机动目标跟踪的一种新的方差自适应滤波算法[J].武汉理工大学学报(交通科学与工程版),2011,35(3):448-452
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |