Abstract
Adjacent obstacles are difficult to be distinguished, and remote obstacles are detected easily to split. Besides, limited deep learning samples easily result in missed detection of obstacles in urban environment. In view of this, a fast and robust detection method is proposed by fusing Double-Layer Region Growth algorithm and Grid-SECOND detector. At first, SECOND detector is improved by replacing voxel grids with 2D grids and adopting multi-dimensional features to detect obstacles, which can reduce the time consumption and ensure the accurate detection of remote obstacles. Then, the first Region Growing algorithm is used to cluster the undetected and non-empty grids, which can detect obstacles outside the training set. At last, the second Region Growing algorithm is used to refine the detection results of obstacles with larger volume and multi-obstacles grids, and complete the obstacle detection. Through testing in our extracted urban dataset and KITTI dataset, it is verified that the proposed method outperforms state-of-the-art methods and can accurately achieve obstacle detection. The average duration of the entire process is about 50ms.
Highlights
Obstacle detection in urban environment has long been the focus of the research on unmanned driving environment perception
Sparsely Embedded Convolutional Detection (SECOND) detector is improved by replacing voxel grids with 2D grids and adopting multi-dimensional features to detect obstacles, which can solve the problem of remote obstacle detection splitting
The second Region Growth algorithm is used to refine the obstacles that are characterized by large volume and may contain multiple targets
Summary
Obstacle detection in urban environment has long been the focus of the research on unmanned driving environment perception. J. Li et al.: Fast Obstacle Detection Method by Fusion of Double-Layer Region Growing Algorithm and Grid-SECOND Detector due to its high detection and classification accuracy. In order to improve the detection accuracy and solve the problems suffered by the above methods, a fast and robust obstacle detection method is proposed in this paper. SECOND detector is improved by replacing voxel grids with 2D grids and adopting multi-dimensional features to detect obstacles, which can solve the problem of remote obstacle detection splitting. DATA PREPROCESSING HDL-64 LIDAR is adopted in this paper to improve the ability of obstacle detection in urban environment, which can generate nearly 130000 points per frame.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
