图像处理中的绝缘子缺陷检测方法

目前国内外生产上钢化玻璃绝缘子的缺陷检测均采用传统的人工肉眼检测方法,难以满足大规模、自动化生产的需要.为此,结合基于形态学的特征检测和基于BP神经网络的缺陷分类检测,提出一种基于图像处理的玻璃件缺陷检测方法.实验结果证明了该方法的有效性.

Defect detection method on insulators by image processing

Defect detection in toughened glass insulators is a main factor to evaluate product quality.The method of traditional manual detections is adopted in industry at home and abroad but can not satisfy the demand of large automatic productions.Therefore,with a morphological feature detection approach and the BP neural network,an approach for defect detection in glass insulators based on machine vision technology is presented.The effectiveness of this method is proved in experiments.