Industry 4.0 is inseparable from intelligent manufacturing, and intelligent manufacturing is inseparable from machine vision. Machine vision is a rapidly developing new technology in recent years that utilizes optical and electromagnetically integration to enable machines to have visual functions. Machine vision has opened a "new window" for industrial automation. In the era of intelligent manufacturing, machine vision has great potential for development.

Machine vision is the use of machines instead of the human eye for measurement and judgment. The application of machine vision is mainly manifested in four aspects:

1. Guidance and positioning: Loading and unloading use machine vision to locate and guide robot arms to accurately grasp.
2. Appearance inspection: Check whether there are quality issues with the products on the production line. This link is also the most important part of replacing labor.
3.  High precision detection: Some products have high accuracy, reaching 0.01~0.02m or even u, which cannot be detected by the human eye and must be completed by machines.
4.  Identification: Data traceability and collection, suitable for other applications such as automotive parts, food, medicine, etc.

Taking detection applications as an example, briefly introduce the working principle of machine vision. Collecting images of the tested equipment using industrial cameras is the most important part of machine vision. In order to detect all the required functions, it is necessary to continuously adjust the parameters of the light source and camera, and then use professional image processing software to convert the analog signal into a digital signal. Then, calculate and extract the detection features of the target, such as color, whether there are scratches on the device surface, whether the size is qualified, whether the surface coating is uniform, etc. Then, output the results and feedback to the mechanical end for device inspection, Selecting unqualified devices involves three processes: image acquisition, information processing, and sorting completion.