Machine Vision-based Defect Detection for Glass Insulators
This study aims to address the core issue of power equipment safety,which is the detection of violation in insu-lators usage,and proposes a technology based on machine vision.The image acquisition unit vertically irradiates the insu-lator with an LED red light source and detects the reflected light with a CCD camera,and collects real-time signals through an image acquisition card.The image processing and defect recognition module uses image processing software such as MATLAB,and undergoes corresponding image refinement preprocessing,threshold segmentation,and feature extraction to effectively detect and classify defects.In terms of image preprocessing,the technology adopts grayscale transformation and image filtering techniques,particularly improving Gaussian low-pass filtering to balance the needs of noise removal and detail protection.Subsequently,the image is divided into different regions through threshold segmenta-tion and key feature parameters are extracted,which helps to accurately identify and classify different defects.The tech-nology is verified through experiments capable of achieving detection of bubble and crack defects.
NETL
NSTL
万方数据
