Impact of color and mixing proportion of synthetic point clouds on semantic segmentation

TL;DR

This study investigates how the color and mixing proportion of synthetic point clouds affect deep learning-based semantic segmentation, revealing that real colors and higher mixing proportions improve model performance.

Contribution

Introduces methods to generate synthetic point clouds with real and uniform colors and provides benchmarks to evaluate their impact on segmentation accuracy.

Findings

Real-colored SPC improves segmentation performance by 8.2%.

Higher than 70% mixing proportion of SPC enhances model accuracy.

SPC can effectively replace real data for training large building element detectors.

Abstract

Deep learning (DL)-based point cloud segmentation is essential for understanding built environment. Despite synthetic point clouds (SPC) having the potential to compensate for data shortage, how synthetic color and mixing proportion impact DL-based segmentation remains a long-standing question. Therefore, this paper addresses this question with extensive experiments by introducing: 1) method to generate SPC with real colors and uniform colors from BIM, and 2) enhanced benchmarks for better performance evaluation. Experiments on DL models including PointNet, PointNet++, and DGCNN show that model performance on SPC with real colors outperforms that on SPC with uniform colors by 8.2 % + on both OA and mIoU. Furthermore, a higher than 70 % mixing proportion of SPC usually leads to better performance. And SPC can replace real ones to train a DL model for detecting large and flat building elements. Overall, this paper unveils the performance-improving mechanism of SPC and brings new insights to boost SPC's value (for building large models for point clouds).