Papers by Keyword: SLAM

Abstract: There are situations like collapsed buildings or inaccessible indoor spaces for humans, when ground robots may be of the most value. Small robots will likely to get into voids and go deeper than the 18-20 feet that a camera on a probe or a borescope can go into. The ground robots would be used to try to understand the internal layout of the structure and to avoid a secondary collapse, for example. In this paper are presented some results on the attempt to create a low cost mapping and guiding system suitable for small robots based on low cost LIDAR (LIght Detection And Ranging) devices. The aim was to create the mapping and guiding system minimizing the costs and maximizing the performances and capabilities.

Abstract: In this paper, we explore a method for posterior elimination for fast computation of the look-ahead Rao-Blackwellised Particle Filtering (Fast la-RBPF) algorithm for the simultaneous localization and mapping (SLAM) problem in the probabilistic robotics framework. In the case when a lot of SLAM states need to be estimated, large posterior states associated with the correct state may be outnumbered by multiple non-zero smaller posteriors. We show that by masking the low posterior weight states with a Gaussian kernel prior to weight selection the accuracy of the la-RBPF SLAM algorithm can be improved. Simulation results reveal that integrated with the proposed method the fast la-RBPF SLAM performance is enhanced over both the existing RBPF SLAM and the unmodified la-RBPF SLAM algorithms.

Abstract: This study uses a wheeled mobile robot (WMR) to explore unknown indoor environment and build up a map of the unknown environment. The robot utilizes laser measurement sensor with a indoor localization system to detect obstacles and identify unknown environment. The localization system provides the position of the robot and is used for map comparison. Fuzzy theory is applied to controller design. The proposed control scheme can control the wheeled mobile robot move along walls and avoid obstacles. The Iterative Closest Point (ICP) and the KD-tree are utilized. With sensed data of obstructions and walls, a map of unknown environment can be generated by curve fitting methods.

Abstract: This paper presents a simultaneous localization, mapping (SLAM) and object tracking (OT) method based on laser and camera data fusion to achieve simultaneous estimation of robot state, environment features states and object’s trajectory in unknown environments. The proposed algorithm is using Full-Correlation Extended Kalman Filter (FCEKF) frameworks, and the system state is combination of robot state, feature states and object state. Object observation is come from two sensors, one is camera and the other is laser. Camshift method is used to get object measurements from camera image, at same time, consistency grid map method is used to get the same object measurements from laser ranger finder, those same object measurements come from different sensor is inputted to FCEKF, and improving robustness and accuracy of system state estimation. The experimental results show that the proposed algorithm is effective to object tracking in outdoor unknown environments.

Abstract: The aims of this paper is to promote object detection and tracking ability of mobile robot in outdoor unknown environments, which based on integration of camera and laser ranger finder data information. First, Camshift method is used to track area of target in image which observed by camera. Second, consistency grid map method is used to detect this moving target based on laser ranger finder observation. Next, two states of the target come from laser and camera sensors are fused to update error of estimation and the new state of target helps to determine search window of Camshift in next loop. Finally, a global optimization method is proposed to improve accuracy of laser-camera calibration. The experimental results show that the proposed algorithm is effective to track object in outdoor unknown environments.

Abstract: This paper presents a novel approach for simultaneous localization, mapping (SLAM) and detection of moving object based on information fusion. We use different information sources, such as laser range scanner, and monocular camera, to improve the ability of distinction between object and background environment. The approach improves the accuracy of SLAM in complex environment, reduces the interference caused by objects, and enhances the practical utility of traditional methods of SLAM. Moreover, the approach expands fields of both research and application of SLAM in combination with target tracking method. Results in real robot experiments show the effectiveness of our method.

Abstract: In this paper, a novel feature-based real-time visual Simultaneous localization and mapping (SLAM) system is proposed. This system generates colored 3-D reconstruction models and 3-D estimated trajectory using a Kinect style camera. Microsoft Kinect, a low priced 3-D camera, is the only sensor we use in our experiment. Kinect style sensors give RGB-D (red-green-blue depth) data which contains 2D image and per-pixel depth information. ORB (Oriented FAST and Rotated BRIEF) is the algorithm used to extract image features for speed up the whole system. Our system can be used to generate 3-D detailed reconstruction models. Furthermore, an estimated 3D trajectory of the sensor is given in this paper. The results of the experiments demonstrate that our system performs robustly and effectively in both getting detailed 3D models and mapping camera trajectory.

Abstract: In this paper we presents a 3D local mapping and localization method with kinect camera. The kinect camera is a low-cost sensor which can provide the information of the RGB image and depth image. The inverse depth parameterization model is introduced to the extended Kalman filter (EKF) to study the location Relations between the camera and feature points. A two-step feature matching method is applied to provide information to SLAM algorithm and synthetize the three-dimensional cloud points. The results of experiments validate the effectiveness of of the proposed local mapping and localization method.

Abstract: Applications on Multi-agent system have been widely studied recently. The positioning of Robotic system is to estimate the position and posture and accurate position estimation. FastSLAM is a SLAM algorithm based on particle filtering, which can perform positioning fast and has been widely applied. This paper applied the genetic particle filtering into SLAM problem to optimize the SLAM algorithm. We present the algorithms based on genetic particle filtering which can obviously reduce the number of particles needed in FastSLAM. The experimental results show that the improvement measures can effectively improve the performance of the algorithm, so that it enables them to maintain a reliable positioning.

Abstract: This paper proposes a non-uniform rational B-spline (NURBS) curve extraction algorithm from 2D laser sensor data and 3D simulated data. In robot localization and mapping application, the raw sensor data cannot be stored due to its need of large storage space. However, only a small number of control points of NURBS curve are needed to be stored to recover the geometrical feature of raw data. To comprise the number of control points and accuracy of the extracted curve, global approximation method is adopted to minimize the error between the raw data and the extracted curve. In extraction process, all the weights are set as one firstly. After find the control points, a weight calculation method is developed to update the weight values. The NURBS curve with new weight has smaller error than with original weights. Finally, NURBS curve extraction results from real 2D laser sensor data and 3D simulated data are shown to check the feasibility of proposed algorithms.