Real-Time Metal-Surface-Defect Detection and Classification Using Advanced Machine Learning Technique
IEEE International Symposium on Broadband Multimedia Systems and Broadcasting, BMSB
In this paper, an advanced machine learning technique is proposed to enable robust real-time metal-surface-detect detection and classification using video streams. The industrial informatics can be inferred from video data according to our proposed new approach. Different from the conventional schemes, our proposed machine-learning technique can detect and classify the metal-surface defects by selecting critical statistical and structural features using Renyi's entropy. To demonstrate the effectiveness of our proposed new detection and classification algorithm, simulation results and performances are compared with the prevalent conventional decision-tree classifier. Based on numerous experimental results, our proposed metal-surface defect detection and classification scheme greatly outperforms the conventional decision-tree classifier.
decision tree, feature selection, Metal-surface defect detection and classification, Renyi's entropy, video data
Applied Data Science
Wei Liu, Kun Yan, Hsiao Chun Wu, Xiangli Zhang, Shih Yu Chang, and Yiyan Wu. "Real-Time Metal-Surface-Defect Detection and Classification Using Advanced Machine Learning Technique" IEEE International Symposium on Broadband Multimedia Systems and Broadcasting, BMSB (2022). https://doi.org/10.1109/BMSB55706.2022.9828748