Machine vision is the study of how computers can achieve vision like humans to achieve "SEE". In fact, vision contains two aspects: "sight" and "perception", that is to say, machine vision not only needs to capture scene information but also needs to understand scene information analysis of color environment. Specifically, it uses cameras and computers to replace the human eyes, so that the machine has similar functions to human to segment, classify, identify, track, discriminate and make decisions on the target. Visual tasks that are very simple for humans can be very complex for machines. In many ways, the ability of the machine vision is not as good as human vision, the reason is that after a great deal of human learning, knowledge and understanding, and have all sorts of things that exist in the real world is the ability to accurately, perfect classified, and the computer is the lack of the corresponding process, just like a baby, it is difficult to distinguish between different people is difficult to identify the object's shape and appearance, expression, etc., But through the interaction with the outside world, learning can gradually master the recognition and understanding ability of things and scenes. Bringing computers up to human vision requires a complete learning process. In addition, the biological eyes have evolved over 500 million years, and the visual system has been perfected continuously, while the camera has only been around for a hundred years. In advanced machine vision problems such as image understanding, computers are generally less able to see than humans. Human and other biological eyes have powerful functions, so the machine vision research process borrowed the functional principles of biological vision, such as Gabor filter frequency and direction expression is similar to the human visual system. The construction of convolutional neural network refers to the way the human brain extracts visual information. Machine vision is to use machines instead of human eyes to measure and judge. Machine vision system refers to the machine vision products (that is, image uptake device, CMO and CCD two) will be ingestion target converted into image signal, transmitted to a special image processing system, according to the pixel distribution and brightness, color and other information, into a digital signal; The image system performs various operations on these signals to extract the features of the target, and then controls the action of the equipment in the field according to the discriminant results.

Machine vision comprehensive application of optics, machinery, electronics, computer hardware and software and other cutting-edge technology, involving computer, image processing, pattern recognition, artificial intelligence, signal processing, opto-mechatronics and other fields. In imaging technology, processing algorithm, and calculate the force platform and industry applications, driven by four core elements of machine vision has been widely used in industry, agriculture, health, security, financial, transportation and other fields, including electronics, electric power, automobile manufacturing, food and beverage, cosmetics, building materials and chemical industry, metal processing, packaging and printing, and many other productive industry and service industry.

With the rise of the concept of "Industry 4.0", domestic machine vision companies have mushroomed. The application of machine vision technology in our country is relatively late, it has been widely used in industrial field for only ten years. At present, the domestic machine vision industry is short of relatively mature automation products, and the market is far from saturated, and there is a large space for development. Domestic enterprises engaged in machine vision research are mainly located in the Pearl River Delta, Yangtze River Delta and Bohai Rim region. It is expected that around 2025, the domestic machine vision industry will enter the mature period of industry. Industry 4.0 is inseparable from intelligent manufacturing, and intelligent manufacturing is inseparable from machine vision technology. Machine vision technology is a necessary means to achieve industrial automation and intelligence, equivalent to the extension of human vision in the machine. Machine vision has a high degree of automation, high efficiency, high precision, and can adapt to the poor environment and other advantages, will play an important role in the realization of industrial automation.

And so on all walks of life, deep into our life, production and work in all aspects, in the information age, playing an increasingly important role. At present, machine vision technology has been widely used in industry, agriculture, national defense, transportation, medical care, finance, sports, entertainment.

In industrial manufacturing testing, the defects on the surface of products are detected by manual testing. With the continuous development of science and technology, especially the development of computer technology, machine vision technology has begun to be widely used in industry. Machine vision technology combined with computer image processing, pattern recognition, comprehensive computer technology, software engineering and other related knowledge in different fields, can accurately and accurately complete the product quality, manual work can not complete the task of detection.