Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection
Three-dimensional object detection technology is an essential component of autonomous driving systems. Existing 3D object detection techniques heavily rely on expensive lidar sensors, leading to increased costs. Recently, the emergence of Pseudo-Lidar point cloud data has addressed this cost issue....
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-09-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/12/18/3920 |
| _version_ | 1797580464464068608 |
|---|---|
| author | Xiangsuo Fan Dachuan Xiao Dengsheng Cai Wentao Ding |
| author_facet | Xiangsuo Fan Dachuan Xiao Dengsheng Cai Wentao Ding |
| author_sort | Xiangsuo Fan |
| collection | DOAJ |
| description | Three-dimensional object detection technology is an essential component of autonomous driving systems. Existing 3D object detection techniques heavily rely on expensive lidar sensors, leading to increased costs. Recently, the emergence of Pseudo-Lidar point cloud data has addressed this cost issue. However, the current methods for generating Pseudo-Lidar point clouds are relatively crude, resulting in suboptimal detection performance. This paper proposes an improved method to generate more accurate Pseudo-Lidar point clouds. The method first enhances the stereo-matching network to improve the accuracy of Pseudo-Lidar point cloud representation. Secondly, it fuses 16-Line real lidar point cloud data to obtain more precise Real Pseudo-Lidar point cloud data. Our method achieves impressive results in the popular KITTI benchmark. Our algorithm achieves an object detection accuracy of 85.5% within a range of 30 m. Additionally, the detection accuracies for pedestrians and cyclists reach 68.6% and 61.6%, respectively. |
| first_indexed | 2024-03-10T22:50:23Z |
| format | Article |
| id | doaj.art-89e2eef9f1bc4dc1bd82e38097a84015 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-10T22:50:23Z |
| publishDate | 2023-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-89e2eef9f1bc4dc1bd82e38097a840152023-11-19T10:23:13ZengMDPI AGElectronics2079-92922023-09-011218392010.3390/electronics12183920Real Pseudo-Lidar Point Cloud Fusion for 3D Object DetectionXiangsuo Fan0Dachuan Xiao1Dengsheng Cai2Wentao Ding3School of Automation, Guangxi University of Science and Technology, Liuzhou 545006, ChinaSchool of Automation, Guangxi University of Science and Technology, Liuzhou 545006, ChinaGuangxi LiuGong Machinery Co., Ltd., Liuzhou 545006, ChinaSchool of Automation, Guangxi University of Science and Technology, Liuzhou 545006, ChinaThree-dimensional object detection technology is an essential component of autonomous driving systems. Existing 3D object detection techniques heavily rely on expensive lidar sensors, leading to increased costs. Recently, the emergence of Pseudo-Lidar point cloud data has addressed this cost issue. However, the current methods for generating Pseudo-Lidar point clouds are relatively crude, resulting in suboptimal detection performance. This paper proposes an improved method to generate more accurate Pseudo-Lidar point clouds. The method first enhances the stereo-matching network to improve the accuracy of Pseudo-Lidar point cloud representation. Secondly, it fuses 16-Line real lidar point cloud data to obtain more precise Real Pseudo-Lidar point cloud data. Our method achieves impressive results in the popular KITTI benchmark. Our algorithm achieves an object detection accuracy of 85.5% within a range of 30 m. Additionally, the detection accuracies for pedestrians and cyclists reach 68.6% and 61.6%, respectively.https://www.mdpi.com/2079-9292/12/18/39203D object detectionstereo matchingPseudo-LidarReal Pseudo-Lidar |
| spellingShingle | Xiangsuo Fan Dachuan Xiao Dengsheng Cai Wentao Ding Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection Electronics 3D object detection stereo matching Pseudo-Lidar Real Pseudo-Lidar |
| title | Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection |
| title_full | Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection |
| title_fullStr | Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection |
| title_full_unstemmed | Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection |
| title_short | Real Pseudo-Lidar Point Cloud Fusion for 3D Object Detection |
| title_sort | real pseudo lidar point cloud fusion for 3d object detection |
| topic | 3D object detection stereo matching Pseudo-Lidar Real Pseudo-Lidar |
| url | https://www.mdpi.com/2079-9292/12/18/3920 |
| work_keys_str_mv | AT xiangsuofan realpseudolidarpointcloudfusionfor3dobjectdetection AT dachuanxiao realpseudolidarpointcloudfusionfor3dobjectdetection AT dengshengcai realpseudolidarpointcloudfusionfor3dobjectdetection AT wentaoding realpseudolidarpointcloudfusionfor3dobjectdetection |