I. Concept of Machine Vision
Machine vision is an interdisciplinary science based on visual skills. Its core technology is visual processing, which is used to perform further detection and control. The Machine Vision Division of the Society of Manufacturing Engineers (SME) and the Automation Vision Division of the Robotic Industries Association (RIA) define machine vision as: "Machine vision is a device that automatically receives and processes images of a real object using optical equipment and non-contact sensors to obtain the necessary information or to control the movement of a robot." A machine vision system is the practical manifestation of machine vision.
II. Types of Machine Vision Systems
According to the definition of machine vision, based on optical characteristics, it can be divided into visible light machine vision systems (based on sensors such as CCD or CMOS) and non-visible light (based on sensors such as X-ray, infrared light, and ultrasound) machine vision systems. Based on the application field, it can be divided into industrial machine vision systems and non-industrial machine vision systems. Generally, people engaged in industrial automation only refer to industrial machine vision systems as machine vision and non-industrial machine vision systems as image equipment or image systems.
III. Characteristics of Machine Vision Systems
■ High precision: A well-designed machine vision system can perform spatial measurement on one of a thousand or more targets. Because this measurement does not require touching the target, there is no damage or risk to the target, and due to the use of computer technology, it has extremely high accuracy.
■ Continuity: Machine vision systems can save people from fatigue. Since there are no human operators, there are no human-induced changes in operation. Multiple systems can be set to operate independently.
■ Flexibility: Machine vision systems can acquire or measure various types of information. When applications change, only corresponding software modifications or upgrades are needed to adapt to the new requirements.
■Standardization: The core of a machine vision system is visual image technology. Therefore, the standards of machine vision system products from different manufacturers are the same, which greatly facilitates the widespread application of machine vision.
IV. Composition of Machine Vision System
Machine vision systems use computers to analyze an image, draw conclusions based on the analysis, and then issue instructions for the next step. Currently, machine vision systems have two main applications:
1. Machine vision systems can explore targets (monitoring, detection, and control).
2. Machine vision can also be used to invent a component, that is, to directly guide the manufacturing process (virtual manufacturing) using a combination of optical instruments and software. Regardless of the application, a typical machine vision system consists of the following subsystems or some of them: sensing and detection system, light source system, optical system, data collection system (camera), image processing system (or image collection card), image measurement and control system (image and control software), monitoring system, communication/input/output system, execution system, alarm system, etc. A machine vision system can be further broken down into product groups:
1) Sensing system: Sensors and their associated sensor controllers, etc.;
2) Light source system: Light source and its associated light source controller, etc.;
3) Optical system: aperture, lens, and optical interface, etc.;
4) Data collection system: digital camera, CCD, CMOS, infrared camera, ultrasonic probe;
5) Image processing system: image collection card, data control card, etc.;
6) Image measurement and control system: software for image acquisition, image processing, image analysis, and automatic control;
7) Monitoring system: monitoring devices;
8) Communication/Input/Output System: Communication links or input/output devices;
9) Actuator: robotic arm and control unit;
10) Alarm system: Alarm equipment and control unit;
11) Machine vision system equipment with special functions based on specific occupational needs.
The products in these product groups that are typical of machine vision system products include: light sources, optics, cameras, data collection cards, measurement and control boards, embedded systems, software, chips, robotic arms, and machine vision system equipment configured according to specific professional applications.
