A survey of machine vision-based monitoring methods for abnormalities in molds

Anomaly monitoring in the mold is a mean to ensure its efficient and stable operation. At present, there are many methods for detecting abnormalities in molds, such as tonnage or strain signal analysis, ultrasonic or magnetostatic detection, and machine vision inspection. This paper compares the advantages and disadvantages of the above methods, lists the existing abnormalities of machine vision mold monitoring methods, and the two major problems (illumination changes and camera vibration) in the process of monitoring solutions are summarized. At the end, the paper summarizes and analyzes the development direction and key points of this field. (paper)