Related Experiment Video
Updated: Jul 11, 2026
10:28
Compact Lens-less Digital Holographic Microscope for MEMS Inspection and Characterization
Published on: July 5, 2016
A binocular machine vision system for three-dimensional surface measurement of small objects
Dimitris Gorpas1, Kostas Politopoulos, Dido Yova
1National Technical University of Athens, School of Electrical and Computer Engineering, Laboratory of Biomedical Optics and Applied Biophysics, 9 Iroon Polutexneiou Street, Zografou Campus, 157 73 Zografou, Greece. dgorpas@mail.ntua.gr
Summary
This study introduces a new 3D volumetric reconstruction method for small objects using binocular machine vision. The system achieves high precision (0.04 mm standard deviation) and robust, automated volumetric measurements.
Area of Science:
- Computer Vision
- Metrology
- 3D Reconstruction
Background:
- Accurate 3D reconstruction of small objects (<1 cm) remains a challenge in computer vision.
- Existing methods lack the precision required for micro-scale applications.
Purpose of the Study:
- To develop a novel volumetric method for precise 3D reconstruction of small objects.
- To improve the accuracy and robustness of 3D imaging systems for micro-scale applications.
Main Methods:
- A binocular machine vision system was employed.
- A novel volumetric approach with crucial z-axis movement was developed.
- Fully automated volumetric algorithms were utilized.
Main Results:
- High precision 3D reconstruction was achieved, with a standard deviation of 0.04 mm.
- The system demonstrated robustness through its lab prototype imaging setup.
- The method requires no further calibration due to automated algorithms and z-axis movement.
Conclusions:
- The developed volumetric method offers a significant advancement in the 3D reconstruction of small objects.
- The system's precision and automated nature make it suitable for various scientific and industrial applications.
- This approach overcomes limitations in current computer vision techniques for micro-scale 3D imaging.
