Development of a Machine Vision Based Non-Contact Metrology System for Inspection of Mechanical Components

Abstract

Machine learning and automation have become an integral part of modern manufacturing industries, enabling faster, smarter, and more reliable production processes. Industrial operations can now be automated using advanced machine learning algorithms combined with high-resolution imaging systems, driving the transition toward Industry 4.0. The central idea of this project is to automate the visual inspection process with high-precision defect detection. One of the major challenges in today’s factories is ensuring consistent quality inspection, as faulty components may still reach customers due to human error. This project focuses on developing an AI-driven system capable of performing dimensional and structural analysis of mechanical components such as pinions, bolts, and nuts. A compact inspection setup is used with a high-resolution camera and precision platform. The captured images are processed using image processing and machine learning techniques to extract parameters and detect defects accurately, ensuring consistency, reliability, and improved productivity.

Introduction

This study presents an automated machine vision-based inspection system designed to improve quality control in manufacturing industries by replacing traditional manual inspection methods.

Manual inspection is often slow, inconsistent, and prone to human error due to fatigue and subjective judgment, making it unsuitable for high-volume production. To overcome these limitations, the proposed system uses computer vision and machine learning techniques to enable fast, accurate, and reliable defect detection in mechanical components.

The system operates using a non-contact inspection setup, where components such as bolts, nuts, and pinions are placed on a backlit platform and captured using a high-resolution camera under controlled lighting conditions. The captured images are then processed through several stages, including grayscale conversion, thresholding, edge detection, and contour analysis, to extract key features such as shape, size, and dimensional parameters. These features are then analyzed using a machine learning model to classify components and identify defects by comparing them with predefined standards.

The results show that the system achieves high precision in dimensional measurement, significantly reduces inspection time, improves consistency, and minimizes human error. It also demonstrates reliable performance for continuous industrial use.

The study concludes that the proposed system enhances productivity and quality assurance in manufacturing by enabling automated, real-time inspection while reducing dependency on manual labor. It supports the transition toward smart manufacturing and Industry 4.0 systems.

Future improvements include the integration of deep learning models for better defect classification, IoT-based remote monitoring, multi-camera systems for complex inspections, and automated sorting mechanisms to reject defective parts instantly.

Conclusion

The proposed system provides an efficient and reliable solution for automated inspection in manufacturing industries. It enhances productivity by enabling fast and continuous inspection of components in real time. The system improves accuracy by eliminating human errors such as fatigue and inconsistency, while ensuring uniform and repeatable inspection results. Additionally, it reduces dependency on manual labor and supports quick identification of defective components, thereby minimizing material wastage and improving overall process efficiency. Overall, the system contributes to better quality control and supports the advancement of smart manufacturing systems.

References

[1] B.Harishyam, M.P. Jenarthanan, R. Rishivanth, Rohit Rajesh, N Sai Girish, “Visual inspection of mechanical components using visual imaging and machine learning”, (Materials Today: Proceedings), 2023. [2] Prof. Vasim Shaikh, Yogeshwar Deshmukh, Prajwal Aute, Girish Pawar, Gunjan Patel, “Automated Quality Inspection Machine Using Computer Vision System”, (IJCRT), Volume 11, Issue 5, May 2023. [3] Hong-Dar Lin, Hsiang-Ling Wu, Chou-Hsien Lin, “A DeepTransfer Learning-Based Visual Inspection System for Assembly Defects in Similar Types of Manual Tool Products”, (MDPI), March 2025. [4] RUI LIU, GUI-JIANG DUAN, “A Framework for Cloud Inspection of Complex Mechanical Parts”, (IEEE), 2023. [5] Priya Charles, Niraj Bhadoria, Simran Gupta, Pooja Satpute, “Survey Paper on Visual Inspection of a Mechanical Part using Machine Learning”, (IJERT), January 2024. [6] Kalaiselvi, Padmanadan, Bhagavan.J., Harishkumar., Kalichandru, “Automatic Work Piece Inspection cum Counter Conveyor”, (IJRASET), May 2025. [7] Huang, “Application of artificial intelligence in mechanical engineering”, (In 2nd International conference on computer engineering, information science & application technology (ICCIA), 2023. [8] Rahmatov, N., Paul, A., Saeed, F., Hong, W. H., Seo, H., Kim, J., “Machine learning–based automated image processing for quality management in industrial Internet of Things”, (International Journal of Distributed Sensor Networks), 2025. [9] P. Arjun, T.T. Mirnalinee, “Machine parts recognition and defect detection in automated assembly systems using computer vision techniques”, (Rev. Téc. Ing. Univ. Zulia.), 2024. [10] Jain, T. Meenu, “Automation and integration of industries through computer vision systems”, (International Journal of Information and Computation Technology), 2025.

Copyright

Copyright © 2026 Prathmesh Shinde, Yash Jadhav, Ayush Jawanjal, Vaibhav Dhole, Vedant Akolkar, Avinash Bhamodre, Uravashi Nale. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.