Research on Classification of Surface Defects of Hot-rolled Steel Strip Based on Deep Learning

Chao WANG, Yu-ting LIU, Ya-ning YANG, Xiang-yu XU, Tao ZHANG

Abstract


Surface defect is one of the important factors affecting the quality of hot-rolled steel strip. Aiming at six typical surface defects of hot-rolled steel strip, a method of surface defect classification based on deep learning is proposed. Based on the Convolutional Neural Networks model and the surface defect data set of Northeast University, the proposed method is verified by experiments. The experimental results show that the recognition rate of surface defects is up to 98.6%, and the detection speed is about 60ms, which meets the requirements of accuracy and speed in industry. The classification and recognition technology of hot strip surface defects proposed in this paper not only has certain theoretical value, but also has practical application prospects.

Keywords


Deep Learning, Surface Defect Detection, Convolutional Neural Networks.


DOI
10.12783/dtcse/ica2019/30756

Full Text:

PDF

Refbacks

  • There are currently no refbacks.