Unsupervised UAV 3D Trajectories Estimation with Sparse Point Clouds

TL;DR

This paper introduces an unsupervised method for detecting small UAVs using fused LiDAR point clouds, achieving accurate real-world tracking without prior training, and demonstrating competitive results in a public challenge.

Contribution

It proposes a novel unsupervised spatial-temporal point cloud processing approach for UAV detection, addressing limitations of traditional detection methods.

Findings

Achieved 4th place in CVPR 2024 UG2+ Challenge

Effective segmentation of foreground UAVs from background in LiDAR data

Open-source code and data for community use

Abstract

Compact UAV systems, while advancing delivery and surveillance, pose significant security challenges due to their small size, which hinders detection by traditional methods. This paper presents a cost-effective, unsupervised UAV detection method using spatial-temporal sequence processing to fuse multiple LiDAR scans for accurate UAV tracking in real-world scenarios. Our approach segments point clouds into foreground and background, analyzes spatial-temporal data, and employs a scoring mechanism to enhance detection accuracy. Tested on a public dataset, our solution placed 4th in the CVPR 2024 UG2+ Challenge, demonstrating its practical effectiveness. We plan to open-source all designs, code, and sample data for the research community github.com/lianghanfang/UnLiDAR-UAV-Est.