Comprehensive Machine Learning Benchmarking for Fringe Projection Profilometry with Photorealistic Synthetic Data
This provides a standardized resource for researchers in optical metrology, but it is incremental as it focuses on benchmarking existing methods rather than introducing new ones.
The paper tackled the lack of datasets and benchmarks in machine learning for fringe projection profilometry by creating a photorealistic synthetic dataset with 15,600 fringe images and 300 depth reconstructions, and found that four neural network architectures achieved similar performance (58-77 mm RMSE) with errors approaching 75-95% of the object depth range.
Machine learning approaches for fringe projection profilometry (FPP) are hindered by the lack of large, diverse datasets and comprehensive benchmarking protocols. This paper introduces the first open-source, photorealistic synthetic dataset for FPP, generated using NVIDIA Isaac Sim with 15,600 fringe images and 300 depth reconstructions across 50 diverse objects. We benchmark four neural network architectures (UNet, Hformer, ResUNet, Pix2Pix) on single-shot depth reconstruction, revealing that all models achieve similar performance (58-77 mm RMSE) despite substantial architectural differences. Our results demonstrate fundamental limitations of direct fringe-to-depth mapping without explicit phase information, with reconstruction errors approaching 75-95\% of the typical object depth range. This resource provides standardized evaluation protocols enabling systematic comparison and development of learning-based FPP approaches.