煤矿救援机器人环境目标图像识别
|
摘要
针对标准尺度不变特征变换(SIFT)算法存在搜索视觉图像中关键点出现计算冗余和目标识别实时性差的问题,提出了一种改进的SIFT算法,并将其应用到煤矿救援机器人的环境信息感知和目标识别匹配中。该方法以马氏距离代替标准SIFT算法中的欧氏距离,简化了特征点提取,避免了特征点的误匹配。现场试验结果表明,改进后的SIFT算法提高了煤矿救援机器人对煤矿井下环境目标识别的实时性和目标匹配的准确性,为煤矿救援自主移动机器人实现避障、行走做好了视觉前提。
In view of problems that traditional scale-invariant feature transform(SIFT)algorithm has calculation redundancy in searching key points of vision images and poor real-time performance in target recognition,an improved SIFT algorithm was proposed which was applied to coal mine rescue robot to realize environmental information perception and target recognition matching. The improved SIFT algorithm adopts Mahalanobis distance to replace Euclidean distance in the traditional SIFT algorithm and simplifies extraction of the image feature points,avoids mismatching of feature point.The field test results show that the improved SIFT algorithm improves real-time performance in underground environmental target recognition and accuracy of target matching of coal mine rescue robot,which makes visual premise for realizing obstacle-avoidance and walking of coal mine rescue autonomous robot.
In view of problems that traditional scale-invariant feature transform(SIFT)algorithm has calculation redundancy in searching key points of vision images and poor real-time performance in target recognition,an improved SIFT algorithm was proposed which was applied to coal mine rescue robot to realize environmental information perception and target recognition matching. The improved SIFT algorithm adopts Mahalanobis distance to replace Euclidean distance in the traditional SIFT algorithm and simplifies extraction of the image feature points,avoids mismatching of feature point.The field test results show that the improved SIFT algorithm improves real-time performance in underground environmental target recognition and accuracy of target matching of coal mine rescue robot,which makes visual premise for realizing obstacle-avoidance and walking of coal mine rescue autonomous robot.
引文
[1]马颂德,张正友.计算机视觉——计算理论与算法基础[M].北京:科学出版社,2003.
[2]SCHMID C,MOHR R.Local gray value invariants for image retrieval[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1997,19(5):530-535.
[3]陈裕,刘庆元.基于SIFT算法和马氏距离的无人机遥感图像配准[J].测绘与空间地理信息,2009,32(6):50-53.CHEN Yu,LIU Qingyuan.The unmanned aerial vehicles remote sensing image registration based on SIFT algorithm and Mahalanobis distanc[J].Geomatics&Spatial Information Technology,2009,32(6):50-53.
[4]LOWE D G.Distinctive image features from scaleinvariant key-points[J].International Journal of Computer Vision,2004,60(2):91-110.
[5]ZHONG Q,ZHAO J.Research on 3Dreconstruction for robot based on SIFT feature[C]//IEEE Advanced Research and Technology in Industry Applications,2014:976-979.
[6]HU M,CHEN J,SHI C.Three-dimensional mapping based on SIFT and RANSAC for mobile robot[C]//IEEE International Conference on Cyber Technology in Automation,Control,and Intelligent Systems,2015:139-144.
[7]ZHU Q,LIU C,CAI C.A novel robot visual homing method based on SIFT features[J].Sensors,2015,15(10):26063-26084.
[8]SATAKE J,CHIBA M,MIURA J.A SIFT-based person identification using a distance-dependent appearance model for a person following robot[C]//IEEE International Conference on Robotics and Biomimetics,2012:962-967.
[9]YANG H,JIANG X,GAO S.Adaptive compressive tracking algorithm based on SIFT features[C]//Third International Conference on Robot,Vision and Signal Processing,2015:78-81.
[10]NAVARRO F,ESCUDERO-VINOLO M,BESCOS J.SP-SIFT:enhancing SIFT discrimination via superpixel-based foreground-background segregation[J].Electronics Letters,2014,50(4):272-274.
[11]JOGLEKAR J,GEDAM S S,MOHAN B K.Image matching using SIFT features and relaxation labeling technique—A constraint initializing method for dense stereo matching[J].IEEE Transactions on Geoscience&Remote Sensing,2014,52(9):5643-5652.
[12]MITTAL A,MURALIDHAR G S,JOYDEEP G,et al.Blind image quality assessment without human training using latent quality factors[J].IEEE Signal Processing Letters,2012,19(2):75-78.
[13]SHEN Xingxing,BAO Wenxing.The remote sensing image matching algorithm based on the normalized cross-correlation and SIFT[J].Journal of the Indian Society of Remote Sensing,2014,42(2):417-422.
[14]黄飞,周军,卢晓东.基于马氏距离的一维距离像识别算法仿真[J].计算机仿真,2010,27(3):31-34.HUANG Fei,ZHOU Jun,LU Xiaodong.The simulation of one-dimensional range profile recognition based on Mahalanobis distance[J].Computer Simulation,2010,27(3):31-34.
[15]郭燕,施喜书,周瑜.煤矿井下照明的分析与研究[J].煤炭工程,2013,12(增刊2):128-129.GUO Yan,SHI Xishu,ZHOU Yu.Analysis and research on underground lighting of coal mine[J].Coal Engineering,2013,12(S2):128-129.
[2]SCHMID C,MOHR R.Local gray value invariants for image retrieval[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,1997,19(5):530-535.
[3]陈裕,刘庆元.基于SIFT算法和马氏距离的无人机遥感图像配准[J].测绘与空间地理信息,2009,32(6):50-53.CHEN Yu,LIU Qingyuan.The unmanned aerial vehicles remote sensing image registration based on SIFT algorithm and Mahalanobis distanc[J].Geomatics&Spatial Information Technology,2009,32(6):50-53.
[4]LOWE D G.Distinctive image features from scaleinvariant key-points[J].International Journal of Computer Vision,2004,60(2):91-110.
[5]ZHONG Q,ZHAO J.Research on 3Dreconstruction for robot based on SIFT feature[C]//IEEE Advanced Research and Technology in Industry Applications,2014:976-979.
[6]HU M,CHEN J,SHI C.Three-dimensional mapping based on SIFT and RANSAC for mobile robot[C]//IEEE International Conference on Cyber Technology in Automation,Control,and Intelligent Systems,2015:139-144.
[7]ZHU Q,LIU C,CAI C.A novel robot visual homing method based on SIFT features[J].Sensors,2015,15(10):26063-26084.
[8]SATAKE J,CHIBA M,MIURA J.A SIFT-based person identification using a distance-dependent appearance model for a person following robot[C]//IEEE International Conference on Robotics and Biomimetics,2012:962-967.
[9]YANG H,JIANG X,GAO S.Adaptive compressive tracking algorithm based on SIFT features[C]//Third International Conference on Robot,Vision and Signal Processing,2015:78-81.
[10]NAVARRO F,ESCUDERO-VINOLO M,BESCOS J.SP-SIFT:enhancing SIFT discrimination via superpixel-based foreground-background segregation[J].Electronics Letters,2014,50(4):272-274.
[11]JOGLEKAR J,GEDAM S S,MOHAN B K.Image matching using SIFT features and relaxation labeling technique—A constraint initializing method for dense stereo matching[J].IEEE Transactions on Geoscience&Remote Sensing,2014,52(9):5643-5652.
[12]MITTAL A,MURALIDHAR G S,JOYDEEP G,et al.Blind image quality assessment without human training using latent quality factors[J].IEEE Signal Processing Letters,2012,19(2):75-78.
[13]SHEN Xingxing,BAO Wenxing.The remote sensing image matching algorithm based on the normalized cross-correlation and SIFT[J].Journal of the Indian Society of Remote Sensing,2014,42(2):417-422.
[14]黄飞,周军,卢晓东.基于马氏距离的一维距离像识别算法仿真[J].计算机仿真,2010,27(3):31-34.HUANG Fei,ZHOU Jun,LU Xiaodong.The simulation of one-dimensional range profile recognition based on Mahalanobis distance[J].Computer Simulation,2010,27(3):31-34.
[15]郭燕,施喜书,周瑜.煤矿井下照明的分析与研究[J].煤炭工程,2013,12(增刊2):128-129.GUO Yan,SHI Xishu,ZHOU Yu.Analysis and research on underground lighting of coal mine[J].Coal Engineering,2013,12(S2):128-129.