Filling Missing Values Matters for Range Image-Based Point Cloud Segmentation

TL;DR

This paper addresses the issue of missing values in range image-based point cloud segmentation by proposing a new projection method and a missing value filling technique, significantly improving segmentation accuracy and speed.

Contribution

It introduces scan unfolding++ (SU++) to reduce missing values and a range-dependent KNN interpolation (KNNI) to fill them, enhancing existing PCS models and achieving state-of-the-art results.

Findings

Improved segmentation performance on SemanticKITTI, SemanticPOSS, and nuScenes datasets.

Proposed methods outperform baseline models in accuracy.

FMVNet and Fast FMVNet achieve superior speed-accuracy trade-offs.

Abstract

Point cloud segmentation (PCS) plays an essential role in robot perception and navigation tasks. To efficiently understand large-scale outdoor point clouds, their range image representation is commonly adopted. This image-like representation is compact and structured, making range image-based PCS models practical. However, undesirable missing values in the range images damage the shapes and patterns of objects. This problem creates difficulty for the models in learning coherent and complete geometric information from the objects. Consequently, the PCS models only achieve inferior performance. Delving deeply into this issue, we find that the use of unreasonable projection approaches and deskewing scans mainly leads to unwanted missing values in the range images. Besides, almost all previous works fail to consider filling in the unexpected missing values in the PCS task. To alleviate this problem, we first propose a new projection method, namely scan unfolding++ (SU++), to avoid massive missing values in the generated range images. Then, we introduce a simple yet effective approach, namely range-dependent $K$-nearest neighbor interpolation ($K$NNI), to further fill in missing values. Finally, we introduce the Filling Missing Values Network (FMVNet) and Fast FMVNet. Extensive experimental results on SemanticKITTI, SemanticPOSS, and nuScenes datasets demonstrate that by employing the proposed SU++ and $K$NNI, existing range image-based PCS models consistently achieve better performance than the baseline models. Besides, both FMVNet and Fast FMVNet achieve state-of-the-art performance in terms of the speed-accuracy trade-off. The proposed methods can be applied to other range image-based tasks and practical applications.