MVM-IOD: An Industrial Object-Centric Benchmark Dataset for the Evaluation of 3D Reconstruction Methods

2026-06-15

AI Summary

This work addresses the challenges of 3D object reconstruction and camera pose estimation in industrial settings, where performance is hindered by object complexity, limited computational resources, and scarcity of real-world data. To bridge this gap, we introduce MVM-IOD, the first multi-view benchmark dataset tailored for industrial objects, comprising RGB images of nine representative workpieces captured under two background conditions using a robotic arm to control the camera along a hemispherical trajectory. The dataset provides ground-truth camera poses and point clouds. Leveraging MVM-IOD, we conduct a systematic evaluation of state-of-the-art methods—including Structure from Motion, Multi-View Stereo, Visual Geometry Grounded Transformer, π³, and 2D Gaussian Splatting—revealing a significant performance drop of feed-forward 3D reconstruction models on out-of-distribution industrial data. Notably, simple preprocessing substantially improves their robustness, offering critical insights for algorithm selection in real-world industrial applications.

Abstract

3D object reconstruction, and camera pose estimation in industrial applications are challenging tasks, as errors are costly while the computation time is often limited. The complexity of typical industrial objects further complicates these tasks. Most of the existing datasets in this context do not depict realistic industrial scenarios. Therefore, we introduce the Machine Vision Metrology Industrial Object Dataset (MVM-IOD). Images of typical industrial objects are captured systematically, by moving a camera, mounted at the end effector of an industrial robot arm, on a hemisphere around the objects. MVM-IOD contains reference camera poses and reference 3D point clouds, the acquired RGB images of 9 objects and 2 background choices resulting in 18 scenes, which allows evaluation of all image based methods that compute a 3D reconstruction, camera poses, or novel views of a scene. Based on MVM-IOD, we extensively evaluate current SOTA 3D reconstruction and camera pose estimation methods, such as Structure from Motion, Multi-View Stereo, recent feed forward methods (Visual Geometry Grounded Transformer, π3), and 2D Gaussian Splatting and report our findings as a baseline for future research. The experiments show that capture setups like ours generate out-of distribution images for feed forward methods, leading to suboptimal point clouds and camera poses. However, these out-of-distribution images can be shifted closer to the training distribution by applying simple preprocessing steps. Consequently, in certain industrial applications, feed forward methods should be used with caution.

Problem

• 3D reconstruction
• camera pose estimation
• industrial applications
• benchmark dataset
• out-of-distribution

Innovation

• industrial 3D reconstruction
• camera pose estimation
• benchmark dataset
• out-of-distribution generalization
• robotic data acquisition