Lightweight framework improves 3d drone location using lidar and images
PRI-Net: A Lightweight Multimodal Framework for 3D UAV Localization
Computer Vision and Pattern Recognition
Summary
Locating drones accurately in 3D space is hard because the data they get from sensors like LiDAR can be sparse and tricky to combine with camera images. The authors propose PRI-Net, a new way to merge these sensor inputs more efficiently. They turn the sparse LiDAR data into denser depth maps, then carefully fuse this with image information while filtering out irrelevant details. This approach helps UAVs find their position better and faster, especially when sending data between the drone and remote servers.
What this means in practice
- •For drone system developers: Enhance drone navigation systems by integrating efficient LiDAR and camera data fusion for more accurate 3D positioning.
- •For autonomous vehicle teams: Improve real-time LiDAR and image fusion techniques for better environmental perception using lightweight models on constrained hardware.
Authors
Zhixuan Chen, Jialiang Lu, Zhong Ye, Yinghui He, Guanding Yu
Abstract
Accurate 3D localization of unmanned aerial vehicles (UAVs) remains challenging for existing multimodal approaches due to sparse LiDAR geometry, modality-imbalanced fusion, and redundant feature transmission over constrained edge-to-server links. To address these limitations, we propose PRI-Net, an efficient and lightweight multimodal fusion framework for UAV localization that integrates point cloud splatting, residual attention fusion, and an information bottleneck. Specifically, a 3D point cloud splatting (3DPCS) strategy is introduced to transform sparse LiDAR observations into geometrically consistent dense depth maps. A residual attention fusion (RAF) module is then designed to alleviate modal bias by using an image branch for coarse estimation and a gated fusion branch for refinement. In addition, a multimodal information bottleneck (MIB) module compacts features by filtering task-irrelevant redundancy. Experiments show that PRI-Net achieves high localization accuracy with lightweight architectures, while reducing feature dimensionality and improving edge-to-server UAV sensing efficiency and robustness.