Laser & Optoelectronics Progress, Volume. 61, Issue 8, 0811005(2024)
Point Cloud 3D Object Detection Based on Improved SECOND Algorithm
Fig. 1. The overall network framework structure of the SECOND algorithm
Fig. 2. Detection results of SECOND algorithm on KITTI validation set. (a) Camera image; (b) point cloud; (c) enlarged detail
Fig. 3. The improved 2D CNN backbone network
Fig. 4. Illustration of the local optimum of smooth L1 loss
Fig. 5. Axis-aligned and rotated bounding boxes
Fig. 6. Bird's eye view of 3D DIoU(2D DIoU schematic)
Fig. 7. The improved multi-task detection head
Fig. 8. Improved SECOND algorithm structure
Fig. 9. Qualitative results of proposed algorithm on KITTI validation set. (a) Camera image; (b) point cloud
Fig. 10. Comparison of qualitative results between the proposed algorithm and the benchmark network (SECOND). (a) Visualization results of SECOND algorithm; (b) visualization results of proposed algorithm
Fig. 11. Comparison of qualitative results of two algorithms. (a) Visualization of SECOND algorithm; (b) visualization of SECOND algorithm+ASFF
Fig. 12. Visualization of spatial weight
|
|
|
|
|
|
Get Citation
Copy Citation Text
Ying Zhang, Liangliang Jiang, Dongbo Zhang, Wanlin Duan, Yue Sun. Point Cloud 3D Object Detection Based on Improved SECOND Algorithm[J]. Laser & Optoelectronics Progress, 2024, 61(8): 0811005
Category: Imaging Systems
Received: Apr. 3, 2023
Accepted: Aug. 2, 2023
Published Online: Mar. 15, 2024
The Author Email: Jiang Liangliang (1017204759@qq.com)