Map Building Method Research Articles

A Fast Multi-Scale of Distributed Batch-Learning Growing Neural Gas for Multi-Camera 3D Environmental Map Building.

Biologically inspired intelligent methods have been applied to various sensing systems in order to extract features from a huge size of raw sensing data. For example, point cloud data can be applied to human activity recognition, multi-person tracking, and suspicious person detection, but a single RGB-D camera is not enough to perform the above tasks. Therefore, this study propose a 3D environmental map-building method integrating point cloud data measured via multiple RGB-D cameras. First, a fast multi-scale of distributed batch-learning growing neural gas (Fast MS-DBL-GNG) is proposed as a topological feature extraction method in order to reduce computational costs because a single RGB-D camera may output 1 million data. Next, random sample consensus (RANSAC) is applied to integrate two sets of point cloud data using topological features. In order to show the effectiveness of the proposed method, Fast MS-DBL-GNG is applied to perform topological mapping from several point cloud data sets measured in different directions with some overlapping areas included in two images. The experimental results show that the proposed method can extract topological features enough to integrate point cloud data sets, and it runs 14 times faster than the previous GNG method with a 23% reduction in the quantization error. Finally, this paper discuss the advantage and disadvantage of the proposed method through numerical comparison with other methods, and explain future works to improve the proposed method.

Research on unmanned vehicle obstacle avoidance technology based on LIDAR and depth camera fusion

Abstract To address the problems of poor accuracy of traditional EKF algorithm in estimating the position of unmanned vehicles and the deficiencies in accuracy and map completeness of the traditional map building method with single-line LiDAR, this paper proposes a method to create fused raster maps realized with multi-source data. Firstly, the combined data of the inertial measurement unit and wheel encoder are corrected by adding the positional information output from the visual odometer using the error-state SLAM algorithm, and the local raster constructed by LiDAR and depth camera is fused frame by frame using the idea of Bayesian estimation to finally generate the fused global map. Then, a four-wheeled mobile unmanned vehicle with a LiDAR sensor and depth camera is selected as the experimental object, and dynamic environment avoidance simulation experiments are conducted to draw conclusions. The simulation experiment results show that when γ = 5.99, the algorithm generates a new local target point p g 2 (17.49, 13.49) and the corresponding getaway path and finally guides the unmanned vehicle to the specified target point, verifying that the method in this paper can achieve the avoidance capability of the unmanned vehicle in the process of getting away from the newly emerged obstacles. This study uses the scanned data of LiDAR for the estimation of the real-time position of the unmanned vehicle to realize obstacle avoidance and path planning of the unmanned vehicle.

Robust Localization for Underground Mining Vehicles: An Application in a Room and Pillar Mine.

Most autonomous navigation systems used in underground mining vehicles such as load-haul-dump (LHD) vehicles and trucks use 2D light detection and ranging (LIDAR) sensors and 2D representations/maps of the environment. In this article, we propose the use of 3D LIDARs and existing 3D simultaneous localization and mapping (SLAM) jointly with 2D mapping methods to produce or update 2D grid maps of underground tunnels that may have significant elevation changes. Existing mapping methods that only use 2D LIDARs are shown to fail to produce accurate 2D grid maps of the environment. These maps can be used for robust localization and navigation in different mine types (e.g., sublevel stoping, block/panel caving, room and pillar), using only 2D LIDAR sensors. The proposed methodology was tested in the Werra Potash Mine located at Philippsthal, Germany, under real operational conditions. The obtained results show that the enhanced 2D map-building method produces a superior mapping performance compared with a 2D map generated without the use of the 3D LIDAR-based mapping solution. The 2D map generated enables robust 2D localization, which was tested during the operation of an autonomous LHD, performing autonomous navigation and autonomous loading over extended periods of time.

Tightly coupled laser-inertial pose estimation and map building based on B-spline curves

Simultaneous localization and mapping (SLAM) plays a key role in 3D environment modeling and mobile robot environment perception. However, the traditional discrete-time Laser-inertial SLAM methods are not robust due to the imbalanced registration steps between a single LiDAR frame and the global map. This paper proposes a tightly coupled laser-inertial pose estimation and map building method that uses B-spline curves to represent continuous-time trajectory and achieve high robustness of the registration steps. To ensure efficiency, the proposed method separates the SLAM task into an odometer module and a mapping module. The odometer module performs a coarse pose estimation, while the mapping module performs a fine one and builds a global map with 3D LiDAR points. B-spline curves are utilized to integrate both IMU measurement constraints and LiDAR point constraints in the proposed mapping module, which can enhance the association of consecutive LiDAR frames in the optimization step. Besides, the explicit expression of the Jacobi matrix derivation for B-spline-based laser residuals is also introduced to furtherly improve the computation efficiency. Both indoor and outdoor experiments are conducted on a self-collected dataset and a public dataset. Experimental results show that the proposed method can achieve superior performance than the baseline method LIO-mapping.

Laser 3D tightly coupled mapping method based on visual information

PurposeSimultaneous localization and map building (SLAM), as a state estimation problem, is a prerequisite for solving the problem of autonomous vehicle motion in unknown environments. Existing algorithms are based on laser or visual odometry; however, the lidar sensing range is small, the amount of data features is small, the camera is vulnerable to external conditions and the localization and map building cannot be performed stably and accurately using a single sensor. This paper aims to propose a laser three dimensions tightly coupled map building method that incorporates visual information, and uses laser point cloud information and image information to complement each other to improve the overall performance of the algorithm.Design/methodology/approachThe visual feature points are first matched at the front end of the method, and the mismatched point pairs are removed using the bidirectional random sample consensus (RANSAC) algorithm. The laser point cloud is then used to obtain its depth information, while the two types of feature points are fed into the pose estimation module for a tightly coupled local bundle adjustment solution using a heuristic simulated annealing algorithm. Finally, the visual bag-of-words model is fused in the laser point cloud information to establish a threshold to construct a loopback framework to further reduce the cumulative drift error of the system over time.FindingsExperiments on publicly available data sets show that the proposed method in this paper can match its real trajectory well. For various scenes, the map can be constructed by using the complementary laser and vision sensors, with high accuracy and robustness. At the same time, the method is verified in a real environment using an autonomous walking acquisition platform, and the system loaded with the method can run well for a long time and take into account the environmental adaptability of multiple scenes.Originality/valueA multi-sensor data tight coupling method is proposed to fuse laser and vision information for optimal solution of the positional attitude. A bidirectional RANSAC algorithm is used for the removal of visual mismatched point pairs. Further, oriented fast and rotated brief feature points are used to build a bag-of-words model and construct a real-time loopback framework to reduce error accumulation. According to the experimental validation results, the accuracy and robustness of the single-sensor SLAM algorithm can be improved.

Multi-Sensor Fusion Tomato Picking Robot Localization and Mapping Research

To address the current problem of unsatisfactory localization and map building of an intelligent picking robot mobile platform in an unstructured picking environment, a multi-sensor fusion localization and map building method is proposed to fuse the visual odometer position estimation results with IMU and wheel odometer. And the pre-fusion processing of wheel odometer data is proposed to solve the problem of large errors of wheel odometers on uneven ground. The multi-sensor fusion scheme is proposed to fuse vision and laser SLAM methods to build a more information-rich point cloud and raster fusion map for the problem of non-closure caused by sparse branches. The experimental results show that the multi-sensor fusion SLAM method can obtain more accurate, richer, and more robust maps.

A non-uniform quadtree map building method including dead-end semantics extraction

To reduce the complexity of large-scene high-resolution maps while using the dead-end information distributed in the unmanned vehicle driving environment, we propose a novel non-uniform quadtree map-building method including dead-end semantic information extraction. By utilizing quadtree data structures, submaps and a positive-order tree depth organization approach, our proposed map can adapt to the large-scale high-resolution requirement and expand more easily to larger environments. To verify the practicality of our proposed map, we have successfully implemented map matching and path planning in real environments. Additionally, we effectively extract the dead-end semantic information that widely distributes in the environment, which can help unmanned vehicles avoid collisions and improve the search efficiency of the planning procedure. We evaluate our method with KITTI datasets, CARLA Simulator, and our self-collected real-world datasets. The experimental results show that our proposed method significantly reduces the complexity of large-scale high-resolution maps, effectively extracts dead-end semantic information, and has good practicality in real environments. The implementation of our method is released here: https://github.com/biter0088/Non-uniform-quadtree-map.

Multistrategy-Based Environment Map Building Using a Mobile Agent

This letter proposes a mobile agent-based environment map building method, which consists of three stages: 1) boundary detection, 2) global scan, and 3) map construction. According to the agent perception model, an improved wall-walking strategy is used to detect the boundary of the unknown environment. The region surrounded by the boundary is divided into meshes, and a point in each mesh is randomly chosen using a simplified internal spiral traversal strategy and used as the target point that the agent needs to reach in order. Based on this, the D <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">*</sup> Lite algorithm is used to control the movement of the agent to reduce the time for a global scan. Finally, the obstacle distribution and the concentration distribution of the pollution source in the unknown environment are generated through an occupancy probability strategy and a radial basis function neural network, respectively. Experimental results show that the proposed method can control the agent to detect the whole unknown environment with improved robustness and computational efficiency so as to construct the environment map completely and accurately.

A Mobile Robot Mapping Method Integrating Lidar and Depth Camera

For mobile robots in uncertain and complex environments, a single sensor obtains a single set of environmental information, creates maps with low accuracy, is easily disturbed by external factors, and easily collides with obstacles in the navigation process. In this paper, we propose a depth camera and LiDAR fusion map building method, which firstly collects data using the LiDAR and depth camera on board the mobile robot, and then preprocesses the collected radar data and depth data and builds a 2D local environment map using RTAB-MAP algorithm. Finally, the local maps are fused using a voxel filter to stitch together a complete global map to complete the map construction. Experimental validation on a ROS-based mobile robot platform shows that the map built using multi-sensor fusion is more complete, presenting more comprehensive environmental information and laying the foundation for subsequent navigation.

Efficient Clustering for Continuous Occupancy Mapping Using a Mixture of Gaussian Processes.

This paper proposes a novel method for occupancy map building using a mixture of Gaussian processes. Gaussian processes have proven to be highly flexible and accurate for a robotic occupancy mapping problem, yet the high computational complexity has been a critical barrier for large-scale applications. We consider clustering the data into small, manageable subsets and applying a mixture of Gaussian processes. One of the problems in clustering is that the number of groups is not known a priori, thus requiring inputs from experts. We propose two efficient clustering methods utilizing (1) a Dirichlet process and (2) geometrical information in the context of occupancy mapping. We will show that the Dirichlet process-based clustering can significantly speed up the training step of the Gaussian process and if geometrical features, such as line features, are available, they can further improve the clustering accuracy. We will provide simulation results, analyze the performance and demonstrate the benefits of the proposed methods.

Study on 2.5D Map Building and Map Merging Method for Rescue Robot Navigation

Study on 2.5D Map Building and Map Merging Method for Rescue Robot Navigation

Visual-inertial fusion positioning and mapping method based on point-line features

Traditional visual Simultaneous Localisation And Mapping (SLAM) technique has the problem of the strong dependence and weak noise immunity in a structured or weak texture environment. This paper proposes a method of simultaneous location and map building based on the information of the visual point, line features and inertial information fusion. Feature points and feature lines are extracted by camera and fused with inertial information. Deep extraction of point and line features is used to obtain scale information. Then optimise the trajectory by bundle adjustment. The evaluations on the public data set of EuRoc show that the Root Mean Square Error (RMSE) is 0.152 m and the method has excellent accuracy, robustness and timeliness.

A Lane-Level Road Marking Map Using a Monocular Camera

The essential requirement for precise localization of a self-driving car is a lane-level map which includes road markings (RMs). Obviously, we can build the lane-level map by running a mobile mapping system (MMS) which is equipped with a high-end 3D LiDAR and a number of high-cost sensors. This approach, however, is highly expensive and ineffective since a single high-end MMS must visit every place for mapping. In this paper, a lane-level RM mapping system using a monocular camera is developed. The developed system can be considered as an alternative to expensive high-end MMS. The developed RM map includes the information of road lanes (RLs) and symbolic road markings (SRMs). First, to build a lane-level RM map, the RMs are segmented at pixel level through the deep learning network. The network is named RMNet. The segmented RMs are then gathered to build a lane-level RM map. Second, the lane-level map is improved through loop-closure detection and graph optimization. To train the RMNet and build a lane-level RM map, a new dataset named SeRM set is developed. The set is a large dataset for lane-level RM mapping and it includes a total of 25157 pixel-wise annotated images and 21000 position labeled images. Finally, the proposed lane-level map building method is applied to SeRM set and its validity is demonstrated through experimentation.

Noise reduction for radio map crowdsourcing building in WLAN indoor localization system

Recent years have witnessed a growing interest in using WLAN fingerprint-based method for indoor localization system because of its cost-effectiveness and availability compared to other localization systems. In order to rapidly deploy WLAN indoor localization system, the crowdsourcing method is applied to alternate the traditional deployment method. In this paper, we proposed a fast radio map building method utilizing the sensors inside the mobile device and the Multidimensional Scaling (MDS) method. The crowdsourcing method collects RSS and sensor data while the user is walking along a straight line and computes the position information using the sensor data. In order to reduce the noise in the location space of the radio map, the short-term Fourier transform (STFT) method is used to detect the usage mode switching to improve the step determination accuracy. When building a radio map, much fewer RSS values are needed using the crowdsourcing method compared to conventional methods, which lends greater influence to noises and erroneous measurements in RSS values. Accordingly, an imprecise radio map is built based on these imprecise RSS values. In order to acquire a smoother radio map and improve the localization accuracy, the MDS method is used to infer an optimal RSS value at each location by exploiting the correlation of RSS values at nearby locations. Experimental results show that the expected goal is achieved by the proposed method.

A neurobiologically inspired mapping and navigating framework for mobile robots

To improve the cognitive ability of the mobile robot in an unknown environment, this paper proposes an episodic memory based self-organizing learning (EM-SOL) framework for robotic experiences learning, cognitive map building and navigation. This EM-SOL framework uses the spatial cells in the hippocampus and entorhinal cortex to path integrate the robotic kinesthetic cues for dead reckoning, and meanwhile extracts the visual cues to activate state neurons for state recognition. By updating the state neurons network, this framework constructs the robotic episodic memory to store these particular experiences created along the exploration process, and achieves the robotic self-organizing learning. The framework based map building method can optimize the cumulative error by introducing the spatial cells phase reset mechanism, and build an episodic-cognitive map that describes not only the topology relationship but also the cognition relationship between these particular experiences in the environment. Furthermore, a combined topological and metric vector navigation method is presented, which can locate the robot relative to its previous experiences in the memory space, anticipate a preferred global path to the target, and guide the mobile robot to navigate to the target accurately. Finally, the proposed EM-SOL framework based methods are evaluated in both the physical environments and the standard KITTI dataset. The results show that the mobile robot can keep on learning to adapt to the changes in an unknown environment, construct valid episodic memory to store these particular learned experiences, build an episodic-cognitive map and execute the target navigation task with high efficiency and accuracy.

Multi-Robot Indoor Environment Map Building Based on Multi-Stage Optimization Method

For a multi-robot system, the accurate global map building based on a local map obtained by a single robot is an essential issue. The map building process is always divided into three stages: single-robot map acquisition, multi-robot map transmission, and multi-robot map merging. Based on the different stages of map building, this paper proposes a multi-staqe optimization (MSO) method to improve the accuracy of the global map. In the map acquisition stage, we windowed the map based on the position of the robot to obtain the local map. Furthermore, we adopted the extended Kalman filter (EKF) to improve the positioning accuracy, thereby enhancing the accuracy of the map acquisition by the single robot. In the map transmission stage, considering the robustness of the multi-robot system in the real environment, we designed a dynamic self-organized communication topology (DSCT) based on the master and slave sketch to ensure the efficiency and accuracy of map transferring. In the map merging stage, multi-layer information filtering (MLIF) was investigated to increase the accuracy of the global map. We performed simulation experiments on the Gazebo platform and compared the result of the proposed method with that of classic map building methods. In addition, the practicability of this method has been verified on the Turtlebot3 burger robot. Experimental results proved that the MSO method improves the accuracy of the global map built by the multi-robot system.

ERASOR: Egocentric Ratio of Pseudo Occupancy-Based Dynamic Object Removal for Static 3D Point Cloud Map Building

Scan data of urban environments often include representations of dynamic objects, such as vehicles, pedestrians, and so forth. However, when it comes to constructing a 3D point cloud map with sequential accumulations of the scan data, the dynamic objects often leave unwanted traces in the map. These traces of dynamic objects act as obstacles and thus impede mobile vehicles from achieving good localization and navigation performances. To tackle the problem, this paper presents a novel static map building method called ERASOR, Egocentric RAtio of pSeudo Occupancy-based dynamic object Removal, which is fast and robust to motion ambiguity. Our approach directs its attention to the nature of most dynamic objects in urban environments being inevitably in contact with the ground. Accordingly, we propose the novel concept called pseudo occupancy to express the occupancy of unit space and then discriminate spaces of varying occupancy. Finally, Region-wise Ground Plane Fitting (R-GPF) is adopted to distinguish static points from dynamic points within the candidate bins that potentially contain dynamic points. As experimentally verified on SemanticKITTI, our proposed method yields promising performance against state-of-the-art methods overcoming the limitations of existing ray tracing-based and visibility-based methods.

Vision‐based map building and path planning method in unmanned air/ground vehicle cooperative systems

To reduce the casualties and improve the efficiency of robot, unmanned air/ground vehicle (UAV/UGV) cooperative systems are proposed. In this study, mathematical algorithms are applied in a typical-rescue scenario. As a ‘flying eye’, the UAV provided the global information by vision sensor. Then, the images were processed with SURF algorithm and image binarisation to model the environment. Based on the UAV's information, the optimised A* algorithm was proposed. Finally, a feasible path was designed for UGV to rescue. Experiments are performed to evaluate the performances of the proposed method. Results show that SURF algorithm and image binarisation can realise the accuracy and robustness of map building. The optimised A* algorithm can provide a real-time and feasible path.

Sparse semantic map building and relocalization for UGV using 3D point clouds in outdoor environments

In this paper, we proposed a sparse semantic map building method and an outdoor relocalization strategy based on this map. Most existing semantic mapping approaches focus on improving semantic understanding of single frames and retain a large amount of environmental data. Instead, we don't want to locate the UGV precisely, but use the imprecise environmental information to determine the general position of UGV in a large-scale environment like human beings. For this purpose, we divide the environment into environment nodes according to the result of scene understanding. The semantic map of the outdoor environment is obtained by generating topological relations between the environment nodes. In the semantic map, only the information of the nodes is saved, so that the storage space can be kept at a very small level with the increasing size of environment. When the UGV receives a new local semantic map, we evaluate the similarity between local map and global map to determine the possible position of the UGV according to the categories of the left and right nodes and the distance between the current position and the nodes. In order to validate the proposed approach, experiments have been conducted in a large-scale outdoor environment with a real UGV. Depending on the semantic map, the UGV can redefine its position from different starting points.

Adapted Error Map Based Mobile Robot UWB Indoor Positioning

Ultrawideband (UWB) has gained a lot of attention in indoor positioning due to its high time resolution, whereas its accuracy could be easily affected by the indoor environment. A UWB error map could represent the distribution of the positioning errors in a static indoor space and help to improve the accuracy of indoor positioning. However, the building of the error map needs much work. In this article, we propose a UWB error map building method, which adapts the distribution of positioning error measurement points to a rough positioning error distribution. An adapted error map based particle filter (AEMBPF), which adopts the error map in the initialization and weight update of the particle set, is then proposed to improve the accuracy of UWB positioning. The experimental results have shown a nearly 50% reduction in the number of measurement points and better positioning accuracy with faster convergence speed.