Research on vision-based single-point strain measurement techniques

Le résumé fourni par la source

Abstract This study develops a vision-based single-point strain measurement system. The hardware consists of a digital microscope, a calibration plate, and a computer, featuring a simple structure and controllable cost, while the software is developed based on Python 3.10 and the OpenCV library. The system integrates image preprocessing, feature extraction, geometric fitting, and data management functions, supports strain calculation for both elliptical and square grids, and offers convenient operation with adaptability to various scenarios. In the inspection of complex product components, combined with AutoForm R8 numerical simulation for targeted measurement in high-risk regions, the average relative error of the major strain is 4.16%. The single-point strain measurement system does not require complex calibration and is simple and efficient to operate while combining low cost and high accuracy, meeting the requirements for local strain measurement of sheet metal components and providing reliable technical support for sheet metal forming quality control with certain engineering application value.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.