Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570108
G. Fritz, C. Seifert, L. Paletta
Autonomous mobile agents require object recognition for high level interpretation and localization in complex scenes. In urban environments, recognition of buildings might play a dominant role in robotic systems that need object based navigation, that take advantage of visual feedback and multimodal information for self-localization, or that enable association to related information from the identified semantics. We present a new method – the informative local features approach – based on an information theoretic saliency measure that is rapidly derived from a local Parzen window density estimation in feature subspace. From the learning of a decision tree based mapping to informative features, it enables attentive access to discriminative information and thereby significantly speeds up the recognition process. This approach is highly robust with respect to severe degrees of partial occlusion, noise, and tolerant to some changes in scale and illumination. We present performance evaluation on our publicly available reference object database (TSG-20) that demonstrates the efficiency of this approach, case wise even outperforming the SIFT feature approach [1]. Building recognition will be advantageous in various application domains, such as, mobile mapping, unmanned vehicle navigation, and systems for car driver assistance.
{"title":"Urban Object Recognition from Informative Local Features","authors":"G. Fritz, C. Seifert, L. Paletta","doi":"10.1109/ROBOT.2005.1570108","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570108","url":null,"abstract":"Autonomous mobile agents require object recognition for high level interpretation and localization in complex scenes. In urban environments, recognition of buildings might play a dominant role in robotic systems that need object based navigation, that take advantage of visual feedback and multimodal information for self-localization, or that enable association to related information from the identified semantics. We present a new method – the informative local features approach – based on an information theoretic saliency measure that is rapidly derived from a local Parzen window density estimation in feature subspace. From the learning of a decision tree based mapping to informative features, it enables attentive access to discriminative information and thereby significantly speeds up the recognition process. This approach is highly robust with respect to severe degrees of partial occlusion, noise, and tolerant to some changes in scale and illumination. We present performance evaluation on our publicly available reference object database (TSG-20) that demonstrates the efficiency of this approach, case wise even outperforming the SIFT feature approach [1]. Building recognition will be advantageous in various application domains, such as, mobile mapping, unmanned vehicle navigation, and systems for car driver assistance.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115540595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570111
M. Rakotondrabe, Y. Haddab, P. Lutz
In this paper, a new type of microsystem is presented : a system able to perform linear and angular motion. First, the microactuator used is studied. An approximation of the working equations is proposed in accordance with Finite-Element simulation. The microactuator used is inserted in the microsystem and a separated modelling of the whole system in linear and in angular motion is given. Stick and skid phases of each motion are independently studied.
{"title":"Step Modelling of a High Precision 2DoF (Linear-Angular) Microsystem","authors":"M. Rakotondrabe, Y. Haddab, P. Lutz","doi":"10.1109/ROBOT.2005.1570111","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570111","url":null,"abstract":"In this paper, a new type of microsystem is presented : a system able to perform linear and angular motion. First, the microactuator used is studied. An approximation of the working equations is proposed in accordance with Finite-Element simulation. The microactuator used is inserted in the microsystem and a separated modelling of the whole system in linear and in angular motion is given. Stick and skid phases of each motion are independently studied.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116032495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570303
B. Selden, Kyu-Jin Cho, H. Asada
This paper describes a new approach to the control of highly nonlinear shape memory alloy (SMA) actuator systems, in which SMA wires are divided into many segments and their thermal states are controlled individually as a group of finite state machines. Instead of driving a current to the entire SMA wire and controlling the wire length based on the analogue strain-temperature characteristics, the new method controls the discrete state (austenite or martensite) of individual segments and thereby controls the total displacement proportional to the number of the austenite segments. Although the inherent property of SMA is highly nonlinear and uncertain with a prominent hysteresis, this Hysteresis Loop Control is robust and stable, providing characteristics similar to a stepping motor. Furthermore, this method can apparently eliminate the latency time associated with phase transition of SMA actuators. Coordination of the multitude of segments having independent thermal states allows for fast response with zero latency time even for thick SMA wires. The new control method is implemented using the Peltier effect thermoelectric devices for selective segment-by-segment heating and cooling. Experiments demonstrate effectiveness of the proposed method, which exploits the inherent hysteresis and nonlinearity of SMA rather than compensating for the nonlinearity.
{"title":"Multi-Segment State Coordination for Reducing Latency Time of Shape Memory Alloy Actuator Systems","authors":"B. Selden, Kyu-Jin Cho, H. Asada","doi":"10.1109/ROBOT.2005.1570303","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570303","url":null,"abstract":"This paper describes a new approach to the control of highly nonlinear shape memory alloy (SMA) actuator systems, in which SMA wires are divided into many segments and their thermal states are controlled individually as a group of finite state machines. Instead of driving a current to the entire SMA wire and controlling the wire length based on the analogue strain-temperature characteristics, the new method controls the discrete state (austenite or martensite) of individual segments and thereby controls the total displacement proportional to the number of the austenite segments. Although the inherent property of SMA is highly nonlinear and uncertain with a prominent hysteresis, this Hysteresis Loop Control is robust and stable, providing characteristics similar to a stepping motor. Furthermore, this method can apparently eliminate the latency time associated with phase transition of SMA actuators. Coordination of the multitude of segments having independent thermal states allows for fast response with zero latency time even for thick SMA wires. The new control method is implemented using the Peltier effect thermoelectric devices for selective segment-by-segment heating and cooling. Experiments demonstrate effectiveness of the proposed method, which exploits the inherent hysteresis and nonlinearity of SMA rather than compensating for the nonlinearity.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116038574","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570592
Yunhui Liu, Hesheng Wang, Kinkwan Lam
This paper presents a new adaptive controller for dynamic image-based visual servoing of a robot manipulator when the intrinsic and extrinsic parameters of the camera are not calibrated. To cope with nonlinear dependence of the image Jacobian on the unknown parameters, this controller employs a matrix called nonscaled image Jacobian which does not depend on the scale factors determined by the depths of feature points. By removing the scale factors, the camera parameters appear linearly in the close-loop dynamics so that a new algorithm, different from Slotine and Li’s, is developed to estimate their values on-line. In the parameter adaptation, in addition to the regressor term, our algorithm also uses the errors between the real and estimated projections of the feature points on the image plane so as to guarantee the convergence of the estimated parameters to the real values up to a scale. A new Lyapunov function is introduced to prove asymptotic convergence of the image errors based on the robot dynamics. Experiments have been conducted to demonstrate the performance of the proposed controller.
{"title":"Dynamic Visual Servoing of Robots in Uncalibrated Environments","authors":"Yunhui Liu, Hesheng Wang, Kinkwan Lam","doi":"10.1109/ROBOT.2005.1570592","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570592","url":null,"abstract":"This paper presents a new adaptive controller for dynamic image-based visual servoing of a robot manipulator when the intrinsic and extrinsic parameters of the camera are not calibrated. To cope with nonlinear dependence of the image Jacobian on the unknown parameters, this controller employs a matrix called nonscaled image Jacobian which does not depend on the scale factors determined by the depths of feature points. By removing the scale factors, the camera parameters appear linearly in the close-loop dynamics so that a new algorithm, different from Slotine and Li’s, is developed to estimate their values on-line. In the parameter adaptation, in addition to the regressor term, our algorithm also uses the errors between the real and estimated projections of the feature points on the image plane so as to guarantee the convergence of the estimated parameters to the real values up to a scale. A new Lyapunov function is introduced to prove asymptotic convergence of the image errors based on the robot dynamics. Experiments have been conducted to demonstrate the performance of the proposed controller.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116063371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570631
R. Vázquez-Martín, J. Lasanta, A. Bandera, F. Hernández
This paper presents a visual attention mechanism for the acquisition of landmarks in an arbitrary scene. The proposed mechanism consists of two consecutive selection stages. The first one employs classical preattentive saliency computa tions to select a reduced set of interest regions from the whole input image (data-driven stage). The second stage selects from the output of the first selection stage the region that can be considered as a potential landmark (model-driven stage). This potential landmark is the input of the attentive stage, that must characterize it and finally determine if this object is a real landmark. The used imaging sensor is a stereo vision system which is capable of providing depth data as well as color images. This stereo vision system is mounted on an autonomous mobile robot and serves map-building and localisation purposes. We present results achieved by applying the proposed visual attention scheme to on-line acquired stereo pairs of indoor and outdoor scenes.
{"title":"Data-and Model-driven Attention Mechanism for Autonomous Visual Landmark Acquisition","authors":"R. Vázquez-Martín, J. Lasanta, A. Bandera, F. Hernández","doi":"10.1109/ROBOT.2005.1570631","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570631","url":null,"abstract":"This paper presents a visual attention mechanism for the acquisition of landmarks in an arbitrary scene. The proposed mechanism consists of two consecutive selection stages. The first one employs classical preattentive saliency computa tions to select a reduced set of interest regions from the whole input image (data-driven stage). The second stage selects from the output of the first selection stage the region that can be considered as a potential landmark (model-driven stage). This potential landmark is the input of the attentive stage, that must characterize it and finally determine if this object is a real landmark. The used imaging sensor is a stereo vision system which is capable of providing depth data as well as color images. This stereo vision system is mounted on an autonomous mobile robot and serves map-building and localisation purposes. We present results achieved by applying the proposed visual attention scheme to on-line acquired stereo pairs of indoor and outdoor scenes.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116205735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570801
Wolfgang Stöcher, G. Biegelbauer
We present enhancements of existing calibration methods that allow the automated calibration of a laser stripe profiler in an industrial environment. The first and most important enhancement is a new design of the calibration object. Basically it is a “wire-frame model” of a cube, where the edges are simply standard aluminium profiles. This allows establishing point correspondences in a simple and robust way without the need of intensity information. Furthermore this hollow kind of a calibration object allows simultaneous multi-view calibration since enough points of six different planes can be seen from many directions. Given a rough calibration (based on rough measurements with ± 50% accuracy) the scanned points can be automatically allocated to the respective planes of the calibration object. Secondly we propose a new way of solving the linear model of a laser stripe profiler. By firstly solving the camera-independent geometry these values can be averaged over the multiple views, which further reduces the local offsets of point clouds and therefore enhances the output of subsequent operations like mesh generation on the combined calibrated point cloud. The paper describes the design of the calibration object and the calibration method in detail and presents experimental results on simulated and real range data sets. The industrial applicability has been demonstrated in a sensor cell with four cameras under real-world conditions including moderate oscillations of the calibration object.
{"title":"Automated Simultaneous Calibration of a Multi-View Laser Stripe Profiler","authors":"Wolfgang Stöcher, G. Biegelbauer","doi":"10.1109/ROBOT.2005.1570801","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570801","url":null,"abstract":"We present enhancements of existing calibration methods that allow the automated calibration of a laser stripe profiler in an industrial environment. The first and most important enhancement is a new design of the calibration object. Basically it is a “wire-frame model” of a cube, where the edges are simply standard aluminium profiles. This allows establishing point correspondences in a simple and robust way without the need of intensity information. Furthermore this hollow kind of a calibration object allows simultaneous multi-view calibration since enough points of six different planes can be seen from many directions. Given a rough calibration (based on rough measurements with ± 50% accuracy) the scanned points can be automatically allocated to the respective planes of the calibration object. Secondly we propose a new way of solving the linear model of a laser stripe profiler. By firstly solving the camera-independent geometry these values can be averaged over the multiple views, which further reduces the local offsets of point clouds and therefore enhances the output of subsequent operations like mesh generation on the combined calibrated point cloud. The paper describes the design of the calibration object and the calibration method in detail and presents experimental results on simulated and real range data sets. The industrial applicability has been demonstrated in a sensor cell with four cameras under real-world conditions including moderate oscillations of the calibration object.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122562208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570644
L. Pimenta, A. R. Fonseca, G. Pereira, R. Mesquita, E. J. Silva, W. Caminhas, M. Campos
This paper addresses the problem of efficiently computing robot navigation functions. Navigation functions are potential functions free of spurious local minima that present an exact solution to the robot motion planning and control problem. Although some methodologies were found in the literature, none of them are easy to implement and generalize for complex shaped workspaces and robots. We discuss some of the difficulties encountered in the current methodologies and propose a novel approach using a Finite Element method for potential field computation.
{"title":"On Computing Complex Navigation Functions","authors":"L. Pimenta, A. R. Fonseca, G. Pereira, R. Mesquita, E. J. Silva, W. Caminhas, M. Campos","doi":"10.1109/ROBOT.2005.1570644","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570644","url":null,"abstract":"This paper addresses the problem of efficiently computing robot navigation functions. Navigation functions are potential functions free of spurious local minima that present an exact solution to the robot motion planning and control problem. Although some methodologies were found in the literature, none of them are easy to implement and generalize for complex shaped workspaces and robots. We discuss some of the difficulties encountered in the current methodologies and propose a novel approach using a Finite Element method for potential field computation.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122722034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570203
G. Paulsen, S. Farritor, T. Huntsberger, H. Aghazarian
The Cliff-bot system consists of three individual planetary rovers that work as a team to explore the surface of a cliff. Two of the rovers, designated “Anchor-bots”, assist the motion of a third rappelling “Cliff-bot” down and along a cliff face using tethers. A decentralized control technique is used to control the motion of the three rovers. The objective of this study is to develop several control algorithms that will create a robust and reliable Cliff-bot system. Control is accomplished by combining and prioritizing several different control algorithms into a hybrid deliberative-reactive control structure. Many different algorithms have been successfully developed and tested to provide the Cliff-bot system with stable and robust navigation of terrain slopes of at least 70 degrees.
{"title":"All Terrain Exploration with the Cliff-bot System","authors":"G. Paulsen, S. Farritor, T. Huntsberger, H. Aghazarian","doi":"10.1109/ROBOT.2005.1570203","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570203","url":null,"abstract":"The Cliff-bot system consists of three individual planetary rovers that work as a team to explore the surface of a cliff. Two of the rovers, designated “Anchor-bots”, assist the motion of a third rappelling “Cliff-bot” down and along a cliff face using tethers. A decentralized control technique is used to control the motion of the three rovers. The objective of this study is to develop several control algorithms that will create a robust and reliable Cliff-bot system. Control is accomplished by combining and prioritizing several different control algorithms into a hybrid deliberative-reactive control structure. Many different algorithms have been successfully developed and tested to provide the Cliff-bot system with stable and robust navigation of terrain slopes of at least 70 degrees.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114402608","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570591
A. Ata, H. Johar
A cubic-splines trajectory planning of a constrained rigid-flexible manipulator to minimize the contact force is considered. Based on the general dynamic model of the chain of flexible links, equations of motion for the manipulator are derived using the extended Hamilton’s principle. An analytical solution for the inverse dynamics problem using the assumed modes method is presented to compute the required joint torques for the tip mass to move along the constrained surface. Cubic-splines interpolation has been adjusted for the joint motion profiles to make sure that the end-effector will follow the prescribed trajectory without degradation. The knot points of the trajectories have been chosen on the constrained surface to achieve minimum contact force as possible.
{"title":"Cubic-Spline Trajectory Planning of a Constrained Flexible Manipulator","authors":"A. Ata, H. Johar","doi":"10.1109/ROBOT.2005.1570591","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570591","url":null,"abstract":"A cubic-splines trajectory planning of a constrained rigid-flexible manipulator to minimize the contact force is considered. Based on the general dynamic model of the chain of flexible links, equations of motion for the manipulator are derived using the extended Hamilton’s principle. An analytical solution for the inverse dynamics problem using the assumed modes method is presented to compute the required joint torques for the tip mass to move along the constrained surface. Cubic-splines interpolation has been adjusted for the joint motion profiles to make sure that the end-effector will follow the prescribed trajectory without degradation. The knot points of the trajectories have been chosen on the constrained surface to achieve minimum contact force as possible.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"116 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122064214","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2005-04-18DOI: 10.1109/ROBOT.2005.1570630
M. Sridharan, Gregory Kuhlmann, P. Stone
Mobile robot localization, the ability of a robot to determine its global position and orientation, continues to be a major research focus in robotics. In most past cases, such localization has been studied on wheeled robots with range finding sensors such as sonar or lasers. In this paper, we consider the more challenging scenario of a legged robot localizing with a limited field-of-view camera as its primary sensory input. We begin with a baseline implementation adapted from the literature that provides a reasonable level of competence, but that exhibits some weaknesses in real-world tests. We propose a series of practical enhancements designed to improve the robot’s sensory and actuator models that enable our robots to achieve a 50% improvement in localization accuracy over the baseline implementation. We go on to demonstrate how the accuracy improvement is even more dramatic when the robot is subjected to large unmodeled movements. These enhancements are each individually straightforward, but together they provide a roadmap for avoiding potential pitfalls when implementing Monte Carlo Localization on vision-based and/or legged robots.
{"title":"Practical Vision-Based Monte Carlo Localization on a Legged Robot","authors":"M. Sridharan, Gregory Kuhlmann, P. Stone","doi":"10.1109/ROBOT.2005.1570630","DOIUrl":"https://doi.org/10.1109/ROBOT.2005.1570630","url":null,"abstract":"Mobile robot localization, the ability of a robot to determine its global position and orientation, continues to be a major research focus in robotics. In most past cases, such localization has been studied on wheeled robots with range finding sensors such as sonar or lasers. In this paper, we consider the more challenging scenario of a legged robot localizing with a limited field-of-view camera as its primary sensory input. We begin with a baseline implementation adapted from the literature that provides a reasonable level of competence, but that exhibits some weaknesses in real-world tests. We propose a series of practical enhancements designed to improve the robot’s sensory and actuator models that enable our robots to achieve a 50% improvement in localization accuracy over the baseline implementation. We go on to demonstrate how the accuracy improvement is even more dramatic when the robot is subjected to large unmodeled movements. These enhancements are each individually straightforward, but together they provide a roadmap for avoiding potential pitfalls when implementing Monte Carlo Localization on vision-based and/or legged robots.","PeriodicalId":350878,"journal":{"name":"Proceedings of the 2005 IEEE International Conference on Robotics and Automation","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122152588","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：