PRISM: Color-Stratified Point Cloud Sampling

2026-01-11

arXiv.org

AI Summary

This work proposes a novel color-aware downsampling method for point clouds that addresses the limitation of traditional approaches, which often disregard color information and thus fail to preserve visually salient features critical for 3D reconstruction. By partitioning the RGB color space into discrete chromatic bins and incorporating a per-bin capacity constraint (k) alongside a chrominance-informed density allocation strategy, the method shifts the sampling paradigm from spatial uniformity to visual complexity–driven selection. This enables preferential retention of points in texture-rich regions while significantly reducing overall point count. Experimental results demonstrate that the proposed approach consistently outperforms conventional techniques—such as random sampling and voxel grid filtering—in preserving perceptually important visual details, thereby yielding superior performance in downstream 3D reconstruction tasks.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Dimensionality Reduction/Feature Selection
• Data Mining & Knowledge Management: Data Compression

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Fairness, privacy, and diversity in economic environments

Abstract

We present PRISM, a novel color-guided stratified sampling method for RGB-LiDAR point clouds. Our approach is motivated by the observation that unique scene features often exhibit chromatic diversity while repetitive, redundant features are homogeneous in color. Conventional downsampling methods (Random Sampling, Voxel Grid, Normal Space Sampling) enforce spatial uniformity while ignoring this photometric content. In contrast, PRISM allocates sampling density proportional to chormatic diversity. By treating RGB color space as the stratification domain and imposing a maximum capacity k per color bin, the method preserves texture-rich regions with high color variation while substantially reducing visually homogeneous surfaces. This shifts the sampling space from spatial coverage to visual complexity to produce sparser point clouds that retain essential features for 3D reconstruction tasks.

Problem

• point cloud sampling
• color diversity
• RGB-LiDAR
• downsampling
• visual complexity

Innovation

• color-stratified sampling
• RGB-LiDAR point clouds
• chromatic diversity
• visual complexity
• point cloud downsampling