Abstract
Object detection in point clouds is an important fundamental problem for many applications such as autonomous driving. The cutting-edge methods generally apply an end-to-end frame to detect the objects from the point clouds, while the traditional methods generally apply a multistep frame that clusters the points before detecting it. The end-to-end frame has been proven to be much better in many aspects, but the good detection effect is supported by a large amount of training data. As is well known, the difficulty of creating point clouds datasets is much greater than that of images datasets. Moreover, when the scenes undergo significant changes, the detection effect of end-to-end frame often decreases significantly. Therefore, on the basis of traditional detection algorithms, we propose a flexible and highly transferable two-stage algorithm AF3D where the point clouds is firstly clustered into clusters which were then detected by a classification network based on deep learning. We verify the algorithm on both the KITTI dataset and our own dataset, and compared it with Pointpillars. The conclusion indicates that AF3D has better robustness than Pointpillars in unfamiliar scenes.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Zamanakos, G., Tsochatzidis, L., Amanatiadis, A., Pratikakis, I.: A comprehensive survey of LIDAR-based 3D object detection methods with deep learning for autonomous driving. Comput. Graph. 99, 153–181 (2021)
Ghasemieh, A., Kashef, R.: 3D object detection for autonomous driving: methods, models, sensors, data, and challenges. Transport. Eng 8, 100115 (2022)
Gupta, A., Anpalagan, A., Guan, L., Khwaja, A.S.: Deep learning for object detection and scene perception in self-driving cars: survey, challenges, and open issues. Array 10, 100057 (2021)
Wang, Z., Li, Q., Zhang, Z., Wang, K., Yang, J.: Research progress of unmanned vehicle point cloud clustering algorithm. World Sci.-Tech. R & D 43, 274–285 (2021)
Wang, S., Liu, C., Xing, S.: Review on k-means clustering algorithm. J. East China Jiaotong Univ. 39, 119–126 (2022)
Ikotun, A.M., Almutari, M.S., Ezugwu, A.E.: K-Means-based nature-inspired metaheuristic algorithms for automatic data clustering problems: recent advances and future directions. Appl. Sci. 11, 11246 (2021)
Xue, L., Qi, C., Zhang, B., Zhang, X., Wu, C.: Object size and orientation recognition based on 3d points cloud Euclideam clustering and RANSAC boundary fitting. Machine Des. Res. 34, 44–48+53 (2018)
Chen, W., Shi, H.: Improved DBSCAN clustering algorithm based on KD tree. Comput. Syst. Appl. 31, 305–310 (2022)
Fan, X., Xu, G., Li, W., Wang, Q., Chang, L.: Target segmentation method for three-dimensional LiDAR point cloud based on depth image. Chin. J. Lasers 46, 292–299 (2019)
Zhang, C., Huang, W., Niu, T., Liu, Z., Li, G., Cao, D.: Review of clustering technology and its application in coordinating vehicle subsystems. Automot. Innov. 6, 89–115 (2023)
Wang, X., Wu, L., Chen, H., Shi, H.: Region segmentation of point cloud data based on improved particle swarm optimization fuzzy clustering. Opt. Precis. Eng. 25, 563–573 (2017)
Shi, S., Wang, X., Li, H.: PointRCNN: 3D object proposal generation and detection from point cloud. In: 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2019). pp. 770–779. IEEE Computer Society, Los Alamitos (2019)
Shi, S., et al.: PV-RCNN: Point-voxel feature set abstraction for 3D object detection. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 10526–10535. IEEE, Seattle, WA, USA (2020)
Li, Z., Wang, F., Wang, N.: LiDAR R-CNN: an efficient and universal 3D object detector. In: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2021, pp. 7542–7551. IEEE Computer Society, Los Alamitos (2021)
Deng, J., Shi, S., Li, P., Zhou, W., Zhang, Y., Li, H.: Voxel R-CNN: towards high performance voxel-based 3d object detection. In: Thirty-Fifth Aaai Conference on Artificial Intelligence, Thirty-Third Conference on Innovative Applications of Artificial Intelligence and the Eleventh Symposium on Educational Advances in Artificial Intelligence, pp. 1201–1209. Assoc. Advancement Artificial Intelligence, Palo Alto (2021)
Qi, C.R., Su, H., Mo, K., Guibas, L.J.: PointNet: deep learning on point sets for 3D classification and segmentation. In: 30th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2017), pp. 77–85. IEEE, New York (2017)
Zhang, H.: Organization and visualization of points cloud data based on octree. J. Taiyuan Normal Univ. Nat. Sci. Edition 10, 128–132 (2011)
Fu, C., Li, G., Song, R., Gao, W., Liu, S.: OctAttention: octree-based large-scale contexts model for point cloud compression. Proc. AAAI Conf. Artif. Intell. 36(1), 625–633 (2022). https://doi.org/10.1609/aaai.v36i1.19942
Zermas, D., Izzat, I., Papanikolopoulos, N.: Fast segmentation of 3D point clouds: a paradigm on LiDAR data for autonomous vehicle applications. In: 2017 IEEE International Conference on Robotics and Automation. ICRA 2017, 29 May 2017–3 June 2017, pp. 5067–5073. Institute of Electrical and Electronics Engineers Inc., Singapore, Singapore (2017)
Geiger, A., Lenz, P., Urtasun, R.: Are we ready for autonomous driving? the KITTI vision benchmark suite. In: 2012 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 3354–3361. IEEE, New York (2012)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Luan, H., Gao, Y., Song, Z., Zhang, C. (2023). A Two-Stage 3D Object Detection Algorithm Based on Deep Learning. In: Lu, H., et al. Image and Graphics . ICIG 2023. Lecture Notes in Computer Science, vol 14358. Springer, Cham. https://doi.org/10.1007/978-3-031-46314-3_30
Download citation
DOI: https://doi.org/10.1007/978-3-031-46314-3_30
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-46313-6
Online ISBN: 978-3-031-46314-3
eBook Packages: Computer ScienceComputer Science (R0)