Development and performance evaluation of stereo and structured light reconstruction systems for dimensional metrology application using augmented reality and distance metrics

Abstract This research presents an augmented reality (AR)-based application for dimensional metrology, enabling quantitative evaluation of manufacturing discrepancies by comparing scanned models with CAD counterparts. The accuracy of 3D metrology devices is influenced by data acquisition conditions and reconstruction techniques. This study evaluates two methods: a custom-built multi-stereo camera-based reconstruction system and a commercially available structured light sensor (Structure Sensor Pro). The AR application, developed using Vuforia’s model target technology, automatically detects objects, overlays CAD models, and visualizes scanned models from both reconstruction methods. It also computes key metrology metrics, including chamfer distance (CD) and Hausdorff distance (HD), to assess deviations. Additionally, three novel distance metrics—density-aware CD, oriented CD, and directional CD—are introduced, further enhancing quantitative evaluation in dimensional metrology. Experimental results demonstrate that the multi-stereo system outperforms the Structure Sensor Pro, achieving 17% lower CD values (0.0079 m vs. 0.0095 m) and 90% lower Earth mover’s distance (0.0054 m vs. 0.058 m). A histogram-based analysis of the HD indicates that 80% of the points in the stereo method exhibit errors below a critical threshold, compared to only 30% in the structured light method, signifying higher reconstruction precision. Measurement uncertainty study further validates the stereo system’s superiority, showing 36% lower error values on average in uncertainty metrics.