基于深度学习的工业零件缺陷检测算法研究

Research on Defect Detection Algorithms for Industrial Parts Based on Deep Learning

随着科学技术的发展,我国工业化水平也在迅速提高,工业零件质量的优劣将直接影响产品的性能,在投入使用前有必要对零件表面进行缺陷检测。笔者分析了传统零件缺陷检测方式的不足,提出一种基于深度学习的零件缺陷检测方法,在原有的BP神经网络算法中融入了基于深度学习的LLENet算法。该方法有效解决了检测效率低、检测速度慢等问题,提高了零件缺陷检测的精度。

With the development of science and technology,the level of industrialization in China is also improving rapidly.The quality of industrial parts will directly affect the performance of products.It is essential to detect the defects of industrial parts.In this paper,the shortcomings of traditional defect detection methods are analyzed,and a defect detection method based on deep learning is proposed.The LLENet algorithm based on deep learning is integrated into the original BP neural network algorithm.This method effectively solves the problems of low detection efficiency and slow detection speed,and improves the accuracy of parts defect detection.