Machine vision is a new technology developed in recent years. It makes use of the means of optical, mechanical and electrical integration to make the machine have the function of vision. By introducing machine vision into the field of inspection, online high-precision and high-speed measurement can be realized in many occasions. At the same time, the theory of machine vision detection technology has been developed step by step.
What is machine vision?
Machine vision, in short, is to use machines instead of human eyes to do measurement and judgment. It mainly uses computer to simulate human visual function, extract information from the image of objective things, process and understand, and finally used for actual detection, measurement and control. The typical mechanism of machine vision consists of five parts: lighting, lens, camera, image acquisition card and vision processor.
The technology derived from this is machine vision technology, which is an interdisciplinary subject involving artificial intelligence, neurobiology, psychophysics, computer science, image processing, pattern recognition and many other fields. Machine vision technology is characterized by high speed, large amount of information and multi-function.
The development of machine vision
The research of machine vision began in the mid-1960s by American scholar L.R. Roberts on understanding the building block world composed of polyhedrons. The techniques used at that time, such as preprocessing, edge detection, contour line construction, object modeling and matching, have been applied in machine vision. In the 1970s, machine vision formed several important research branches: ① image processing of target guidance; ② parallel algorithm of image processing and analysis; ③ extraction of 3D information from 2D images; ④ analysis of sequence images and evaluation of motion parameters; ⑤ representation of visual knowledge; ⑥ knowledge base of visual system.
What are the difficulties of machine vision technology design?
First: the stability of lighting
Industrial vision applications are generally divided into four categories: positioning, measurement, detection and recognition. Among them, the measurement requires the highest stability of illumination. As long as the illumination changes by 10-20%, the measurement result may deviate by 1-2 pixels. This is not a software problem, it is illumination variation, which leads to the change of the edge position on the image, and even the most powerful software can not solve it In order to solve the problem, we must eliminate the interference of ambient light from the system design point of view, and ensure the luminous stability of active lighting source. Of course, the resolution improvement of hardware camera is also a way to improve the accuracy and resist environmental interference. For example, the corresponding space size of the previous camera is 10um per pixel, but it is changed to 5um by increasing the resolution. The accuracy can be approximately doubled, and the interference to the environment is naturally enhanced.
Second, the inconsistency of workpiece position
Generally, the first step of the measurement project, whether it is offline detection or online detection, as long as it is fully automated detection equipment, the first step is to find the target to be measured. Every time the object to be measured appears in the shooting field of view, it is necessary to know exactly where the object to be measured is. Even if you use some mechanical fixtures, it can not ensure that the target to be measured appears in the same position every time. This requires the use of positioning function. If the positioning is not accurate, the position of the measurement tool may be inaccurate, and the measurement results may be relatively high Large deviation.
Third: calibration
In general, the following calibrations are needed in high-precision measurement: 1) optical distortion calibration (if you are not using a software lens, it is generally necessary to calibrate), and (2) the projection distortion calibration, that is, the image distortion correction represented by the installation position error, and the calibration of three object image space, that is to calculate the corresponding object space size of each pixel.
However, the current calibration algorithms are based on plane calibration. If the physics to be measured is not planar, some special algorithms will be needed for calibration, which cannot be solved by common calibration algorithms.
In addition, some calibration methods must be designed because it is not convenient to use the calibration board, so the calibration may not be solved by the existing calibration algorithm in the software.
Fourth: the speed of the object
If the object to be measured is not still, but in motion, the image accuracy of motion blur must be considered (blur pixel = object motion speed * camera exposure time), which is not solved by software.
Fifthly, the measurement accuracy of the software
In the measurement application, the accuracy of the software can only be considered as 1 / 2-1 / 4 pixels, and it is better to use 1 / 2 instead of 1 / 10-1 / 30 pixels as positioning applications, because the software can extract very few feature points from the image.
Development and application in China
Experts believe that the early development of machine vision, mainly concentrated in Europe and the United States and Japan. With the shift of global manufacturing center to China, Chinese machine vision market is becoming an important target market for International Machine Vision manufacturers after North America, Europe and Japan. In China, the application of machine vision originated from the introduction of technology in the 1980s. Semiconductor and electronic industry is one of the industries with earlier application of machine vision, most of which are concentrated in PCB printed circuit assembly, component manufacturing, semiconductor and integrated circuit equipment. The application and promotion of machine vision in this industry plays an important role in improving the quality and production efficiency of electronic products effect.
At present, China is becoming one of the most active areas of the world machine vision development, covering almost all sectors of the national economy, including: industry, agriculture, medicine, military, aerospace, meteorology, astronomy, public security, transportation, security, scientific research and other fields. The industrial field is the field with the largest proportion of machine vision application. The important reason is that China has become the processing center of global manufacturing industry. With the high demand of parts processing and its corresponding advanced production line, many machine vision systems and application experiences with international advanced level have also entered China.
The following is a brief introduction to several applications:
1. Food safety monitoring
In the process of production and product quality inspection, it is sometimes necessary for the staff to observe, identify and discover the errors and omissions in the production process. No matter how strong a person's sense of responsibility and attention is, he may be tired, negligent and distracted, causing defective products to flow to the market.
Application of machine vision in food inspection
2. Manufacturing
Manufacturing industry competition intensifies, cost pressure forces it to pay attention to production efficiency, quality will promote the application of machine vision technology. In order to improve the production efficiency and reduce the labor cost, some manual links in industrial production and management are gradually replaced by machines. The characteristic of machine vision system is to improve the flexibility and automation of production. In some dangerous working environment which is not suitable for manual work or where artificial vision is difficult to meet the requirements, machine vision is often used to replace artificial vision; at the same time, in the process of large-scale industrial production, the efficiency of product quality inspection by artificial vision is low and the accuracy is not high, and the detection method of machine vision can greatly improve the production efficiency and the degree of automation of production. Moreover, machine vision is easy to realize information integration, which is the basic technology of computer integrated manufacturing.
At the same time, machine vision technology can also play a role in super standard emission of smoke, sewage and so on. Using machine vision, we can find the fire and smoke in the machine room and workshop in time. The use of face detection and face recognition technology in machine vision can help enterprises to strengthen the control and management of the entrance and exit, improve the management level and reduce the management cost.
3. Solar energy, traffic monitoring
In recent years, the development of emerging industries has brought new market space to machine vision market. In the field of solar energy, solar cell and module manufacturers use machine vision to detect products, identify and track products and assemble products. In the field of traffic monitoring, license plate recognition technology and image analysis technology can be used to automatically identify license plates, find illegal parking, retrograde, and find traffic accident vehicles. In addition, there is a huge space for machine vision technology in the fields of earthquake prevention, landslide, debris flow, volcanic eruption detection, hydrological monitoring and river hydrological condition observation.
Future market prospects
Traditional manufacturing industry is facing new subversion, transformation and upgrading will bring huge market opportunities to China's automation industry. Machine vision, as a highly intelligent product in the automation industry, has great development potential in the future.
In the past few years, China's electronic manufacturers and OEM manufacturers are purchasing a large number of automation equipment to replace manual labor in response to the growing shortage of labor in China, which will reach a climax in the next few years. Taiwan funded factories have chosen to improve the degree of automation, and the upsurge of their automation will come in the next 2-3 years, which will bring new growth point for the application of machine vision products in this industry.
According to an authoritative industry forecast report, the market size of China's machine vision industry will continue to grow, reaching 3 billion yuan in 2015, 3.8 billion yuan in 2016, and 5 billion US dollars by 2018. The new business opportunities brought by the global machine vision market have become the focus of the industry manufacturers.
This article comes from the Internet. It is reprinted for the purpose of spreading knowledge, beneficial learning and research. It is provided free of charge for netizens. It is also used to indicate the author and source. If there is any objection from the copyright owner or publisher, this website will delete it immediately. If you have any questions about the reprint of the article, please let us know so that we can correct it in time.
EN
