What is machine vision? What are its advantages?

 What is machine vision?
Machine vision completes complex industrial tasks reliably and consistently.
According to the Automated Imaging Association (AIA), machine vision covers all industrial and non-industrial applications. It uses a combination of hardware and software capabilities to provide operational guidance for equipment based on image acquisition and processing. . Although industrial computer vision uses many of the same computer vision algorithms and methods as academic/educational and government/military applications, it is different in some ways.
Industrial vision systems require greater ruggedness, reliability, and stability than academic/educational vision systems, and are typically much lower cost than in government/military applications. Therefore, industrial machine vision means low cost, acceptable accuracy, high ruggedness, high reliability, high mechanical properties, and temperature stability.
Machine vision systems rely on protected digital sensors and specialized optics within industrial cameras to capture images, enabling computer hardware and software to process, analyze and measure a variety of properties to aid decision-making.
Take the filling level detection system in the beverage industry as an example (Figure 1). Each bottle of beer passes through an inspection sensor, triggering a vision system that turns on a flash and takes a picture of the bottle. After the image is captured and saved to memory, the vision software processes or analyzes it and issues a pass or fail response based on how filled the bottle is. If the system detects an incorrectly filled bottle (failure condition), a command is issued to the diverter to reject the bottle. The operator can view statistics on rejected bottles and ongoing processing on the display.
Machine vision systems can also perform objective measurements, such as determining spark plug gaps or providing positional information to guide robots in aligning parts during manufacturing. Figure 2 shows how a machine vision system can be used to detect pass or fail on an oil filter (right) and measure the width of the center flap on the bracket (left).

Advantages of machine vision:

Vision can improve quality and productivity while reducing production costs.
Human vision is suitable for qualitative interpretation of complex, unstructured scenes, while machine vision is better at quantitatively determining structured scenes due to its excellent speed, accuracy and repeatability. For example, on a production line, machine vision systems can inspect hundreds or even thousands of parts per minute. Machine vision made with the right camera resolution and optics can easily detect object details that are difficult for the human eye to see.
In terms of avoiding physical contact between the test system and the part under test, machine vision can avoid part damage and avoid maintenance and cost expenses caused by wear and tear of mechanical components. Machine vision can reduce human intervention in the manufacturing process, thereby increasing safety and ease of operation. In addition, it can also avoid artificial contamination of the clean room and protect workers from accidentally entering dangerous environments.
Machine vision helps achieve specific target strategic
goals Machine vision applications
high quality Inspection, measurement, metrology and assembly verification
increase production efficiency Repetitive tasks previously done manually are now done by machine vision systems
Production flexibility Measurement and metrology/robot guidance /Pre-operation verification
reduces downtime and setup time Programming ahead of time to complete product changeovers
More complete information, tighter process control Now manual tasks can provide computer data feedback
Reduce capital equipment costs Adding vision to machines improves their performance and avoids obsolescence Reduced product costs by being eliminated
One vision system gets the job done, eliminating the need for multiple people in the process Multiple early defect detection
reduces scrap rates Inspection, measurement and metrology
inventory management Optical feature recognition and discrimination
reduces floor space No operators are required with vision systems