The edge extraction method of casting images based on improved Canny operator

• Published in: 2022 IEEE International Conference on Mechatronics and Automation (ICMA) pp. 1708 - 1713
• Main Authors: Gao, Qiang, Chen, Jiaqi, Ji, Yuehui, Liu, Junjie, Song, Yu, Wu, Haiqing
• Format: Conference Proceeding
• Language: English
• Published: IEEE 07.08.2022
• Subjects: Canny operator; Casting; Filtering; Image edge detection; Image processing; Liquids; Metal liquid level; Metals; Noise reduction; Quantitative pouring; Shape
• ISSN: 2152-744X
• DOI: 10.1109/ICMA54519.2022.9855947

During the quantitative pouring process, the molten metal temperature is high, and the working environment on site is harsh. Ordinary sensors can not detect the melting height of metal in real-time, so only semi-automatic pouring. This paper proposes an automated pouring image processing technology based on machine vision and focuses on the design steps of machine vision for the acquisition of metal liquid level height. Firstly, feature analysis was carried out according to the shape of the sprue cup, and the image of the metal liquid level in the pouring process was collected in real-time. The original image is processed by binarization, image noise reduction, and liquid column segmentation. In the process of noise reduction, median filtering is used to improve the image edge blur caused by Gaussian filtering. After that, the Otsu algorithm is used to improve the selection of the Canny operator threshold, and the improved Canny operator is used to detect the edge of the image. Finally, edge information is fitted by the least square method. The experimental results show that the proposed edge extraction method has high precision and good adaptability and can effectively solve the problem of metal liquid level detection in pouring images.