LiDAR-Based High-Accuracy Parking Slot Search, Detection, and Tracking 2020-01-5168
The accuracy of parking slot detection is a challenge for the safety of the Automated Valet Parking (AVP), while traditional methods of range sensor-based parking slot detection have mostly focused on the detection rate in a scenario, where the ego-vehicle must pass by the slot. This paper uses three-dimensional Light Detection And Ranging (3D LiDAR) to efficiently search parking slots around without passing by them and highlights the accuracy of detecting and tracking. For this purpose, a universal process of 3D LiDAR-based high-accuracy slot perception is proposed in this paper. First, the method Minimum Spanning Tree (MST) is applied to sort obstacles, and Separating Axis Theorem (SAT) are applied to the bounding boxes of obstacles in the bird’s-eye view, to find a free space between two adjacent obstacles. These bounding boxes are obtained by using common point cloud processing methods. Then the fuzzy analysis method is applied to distinguish the availability and type of free space. Second, a new fitting criterion (minCD), which is robust to the disturbance of rearview mirrors and unevenly distributed points, is proposed to acquire a more accurate contour of the parking slot found in the first step. The fitting results, which contain both the slot posture and the relative slot position, were tracked in a Kalman Filter (KF) until they are accurate enough. Finally, after the parking procedure has started, the corner points of the slot were tracked in Extended Kalman Filter (EKF), in which detection results and dead reckoning (DR) information were fused, to acquire the relative position of the ego-vehicle and the parking slot. During this process, given LiDAR’s blind spots and visual field occlusion, it is necessary to select the most reliable corner as a positioning reference in each step to update the relative position of the ego-vehicle and slot. The effectiveness and accuracy of this approach have been demonstrated in the united simulation environment of Prescan and Simulink.