The American society of manufacturing engineers (
Machine vision branch and the robot industry association (
Automated visual branch of machine vision is defined as: 'machine vision is through the optical device and non-contact sensor automatically receiving and processing of a real image of the object, in order to obtain the required information or device used to control the robot movement'.
In the modern industrial automation production, ranged over a variety of inspection, production monitoring and parts identification application, such as the dimension check spare parts processing, automatic assembly of integrity checks, electronic components automatically locate on an assembly line, the character recognition on the IC, etc.
Usually the human eye can't do the continuous and steady work with high repeatability and intelligent, other physical quantities has sensors to have nowhere to go.
Thus people began to consider using photoelectric imaging system acquisition is accused of target image, and then by the computer or dedicated digital image processing module of processing, according to the image pixel distribution, brightness and color information, such as for size, shape, color and so on.
So, I put the quickness of computers, repeatability, and the height of the human eye visual intelligent and abstract ability, combining the concept of machine vision.
Machine vision system schematic diagram of a successful machine vision system is a through detailed engineering processing to meet the requirements of a series of specific systems.
When these requirements are completely determined, the system is designed and built to meet the requirements of the precision.
The advantages of machine vision include the following: 1, high precision as a precision measuring instruments, design a good vision system can of one thousand or more parts of a spatial measurement.
Because this kind of measurement does not need to contact, so is not wear and dangerous for vulnerable parts.
2, continuity of vision system is to protect people from the fatigue.
Because there is no manual to identify with the naked eye, there would be no operation of man-made errors.
Multiple systems can be set running alone.
3, high cost efficiency with computer's processor prices fell sharply, cost efficiency machine vision system is becoming more and more is also high.
In Europe and the United States, a $10000 worth of visual system can be easily replaced three artificial detection, and each surveyor need the salary of $20000 a year.
In China, the application of exaggerated is one machine vision equipment, test measurement can substitute for hundreds of thousands of people.
In addition, the visual system of operating and maintaining cost is very low.
4, flexibility, vision systems that can perform a variety of different measurements.
When application change, need to corresponding change or upgrade software made to adapt to the new requirements.
Machine vision system has better adaptability than optical or machine sensor.
They make the automatic machine with the diversity, flexibility and restructuring.
When need to change the production process, for machine vision 'tool change' is only the transformation of software rather than replace the expensive hardware.
When the production line after the restructuring, the visual system often can be reused.