Visual Navigation and Path Tracking Using Street Geometry Information for Image Alignment and Servoing

Single camera-based navigation systems need information from other sensors or from the work environment to produce reliable and accurate position measurements. Providing such trustable, accurate, and available information in the environment is very important. The work highlights that the availability of well-described streets in urban environments can be exploited by drones for navigation and path tracking purposes, thus benefitting from such structures is not limited to only automated driving cars. While the drone position is continuously computed using visual odometry, scene matching is used to correct the position drift depending on some landmarks. The drone path is defined by several waypoints, and landmarks centralized by those waypoints are carefully chosen in the street intersections. The known streets’ geometry and dimensions are used to estimate the image scale and orientation which are necessary for images alignment, to compensate for the visual odometry drift, and to pass closer to the landmark center by the visual servoing process. Probabilistic Hough transform is used to detect and extract the street borders. The system is realized in a simulation environment consisting of the Robot Operating System ROS, 3D dynamic simulator Gazebo, and IRIS drone model. The results prove the suggested system efficiency with a 1.4 m position RMS error.

Using Virtual Scanning to Find Optimal Configuration of a 3D Scanner Turntable for Scanning of Mechanical Parts

The article describes a method of simulated 3D scanning of triangle meshes based on ray casting which is used to find the optimal configuration of a real 3D scanner turntable. The configuration include the number of scanners, their elevation above the rotary table and the number of required rotation steps. The evaluation is based on the percentage of the part surface covered by the resulting point cloud, which determines the ability to capture all details of the shape. Principal component analysis is used as a secondary criterion to also evaluate the ability to capture the overall general proportions of the model.

Composite clustering normal distribution transform algorithm

Scan registration is a fundamental step for the simultaneous localization and mapping of mobile robot. The accuracy of scan registration is critical for the quality of mapping and the accuracy of robot navigation. During all of the scan registration methods, normal distribution transform is an efficient and wild-using one. But normal distribution transform will lead to the unreasonable interruption when splitting the grid and can’t express the points’ local geometric feature by prefixed grid. In this article, we propose a novel method, composite clustering normal distribution transform, which comprises the density-based clustering and k-means clustering to aggregate the points with similar local distributing feature. It takes singular value decomposition to judge the suitable degree of one cluster for further division. Meanwhile, to avoid the radiating phenomenon of LIDAR in measuring the points’ distance, we propose a method based on trigonometric to measure the internal distance. The clustering method in composite clustering normal distribution transform could ensure the expression of LIDAR’s local distribution and matching accuracy. The experimental result demonstrates that our method is more accurate and more stable than the normal distribution transform and iterative closest point methods.

Extended Line Map-Based Precise Vehicle Localization Using 3D LIDAR

An Extended Line Map (ELM)-based precise vehicle localization method is proposed in this paper, and is implemented using 3D Light Detection and Ranging (LIDAR). A binary occupancy grid map in which grids for road marking or vertical structures have a value of 1 and the rest have a value of 0 was created using the reflectivity and distance data of the 3D LIDAR. From the map, lines were detected using a Hough transform. After the detected lines were converted into the node and link forms, they were stored as a map. This map is called an extended line map, of which data size is extremely small (134 KB/km). The ELM-based localization is performed through correlation matching. The ELM is converted back into an occupancy grid map and matched to the map generated using the current 3D LIDAR. In this instance, a Fast Fourier Transform (FFT) was applied as the correlation matching method, and the matching time was approximately 78 ms (based on MATLAB). The experiment was carried out in the Gangnam area of Seoul, South Korea. The traveling distance was approximately 4.2 km, and the maximum traveling speed was approximately 80 km/h. As a result of localization, the root mean square (RMS) position errors for the lateral and longitudinal directions were 0.136 m and 0.223 m, respectively.