Today, with the advent of Industry 4.0, machine vision technology is gradually playing a very important role in industrial automation. The continuous innovation of machine vision technology has promoted the progress of industries such as industrial automation, smart security and artificial intelligence. The development of machine vision technology has also brought more development potential and opportunities to the fields in which this technology can be applied.
It's widely acknowledged that humans perceive 80% of their external information through their eyes, with images containing the most information. The emergence of machine vision technology essentially equips machines with "eyes" to perceive the world, enabling them to perform various detection, judgment, recognition, and measurement functions, much like humans do.
Today, let's experience the charm of machine vision technology.
Five typical architectures of machine vision
1.Lighting
I remember attending a machine vision technology conference where a tech expert specifically emphasized the importance of lighting in machine vision. I understand that lighting is a crucial factor affecting the input to a machine vision system, directly impacting the quality of the input data and the effectiveness of the application.
Light sources can be divided into visible light and invisible light. Commonly used visible light sources include incandescent lamps, fluorescent lamps, mercury lamps, and sodium lamps. A drawback of visible light is its instability; how to maintain light stability to a certain extent is a pressing problem to be solved in practice. On the other hand, ambient light can affect image quality, so using a protective screen can reduce its impact.
Illumination systems can be categorized by their illumination methods, including backlighting, front lighting, structured light, and stroboscopic lighting. Backlighting places the object under test between the light source and the camera, offering the advantage of high-contrast images. Front lighting places the light source and camera on the same side of the object, facilitating installation. Structured light illumination projects a grating or line light source onto the object, demodulating its three-dimensional information based on the resulting distortions. Stroboscopic lighting illuminates the object with high-frequency light pulses, requiring the camera to be synchronized with the light source.
2. Lens
The lens is analogous to the human eyeball; in a machine vision system, it is primarily responsible for beam modulation and signal transmission. It is understood that most lenses on the market today can meet the needs of machine vision applications, but more specialized machine vision systems may require custom lenses and coatings. Fortunately, many lens manufacturers with in-house production capabilities are ready to customize lenses to meet these application requirements. Of course, these custom lenses are more expensive, so they are typically only used in specific imaging systems where cost is not a major concern (such as military applications) or in production lines for high-volume consumer products.
3. Industrial cameras
The most fundamental function of industrial cameras in machine vision systems is to convert light signals into electrical signals. Compared to ordinary cameras, they offer higher transmission capabilities, stronger anti-interference capabilities, and more stable imaging. They can be classified in several ways according to different standards: based on the output signal method, they can be divided into analog industrial cameras and digital industrial cameras; based on the chip type, they can be divided into CCD industrial cameras and CMOS industrial cameras.
4. Image acquisition card
Although the image acquisition card is only one component of a complete machine vision system, it plays a crucial role, directly determining the camera's interface: monochrome, color, analog, digital, etc. Typical examples include PCI acquisition cards, 1394 acquisition cards, VGA acquisition cards, and GigE gigabit network acquisition cards. Some of these cards have built-in multiplexers, allowing connection to multiple cameras and simultaneous capture of multiple data streams.
5. Machine vision software
Machine vision software is a key component of automated processing in machine vision systems. Depending on specific application requirements, the software package can be further developed to automatically complete image acquisition, display, storage, and processing. Initially, images are transmitted to dedicated image processing software, where they are converted into digital signals based on pixel distribution, brightness, color, and other information. The machine vision software then performs various calculations on these signals to extract target features, and subsequently controls the actions of on-site equipment based on the judgment results.
Five common application areas
Machine vision can be considered the soul of industrial automation systems. From object/barcode identification, product inspection, and appearance and size measurement to the positioning of robotic arms/transmission equipment, machine vision technology can play a significant role. Therefore, its application range is very wide, and the number of industry applications is dazzling.
1. Image recognition applications
Image recognition utilizes machine vision to process, analyze, and understand images in order to identify targets and objects of various patterns. The most typical application of image recognition in the industrial field of machine vision is QR code recognition, the most common type of barcode we see every day. Large amounts of data are stored in these small QR codes, enabling product tracking and management. Machine vision systems can easily recognize and read barcodes on various material surfaces, greatly improving the efficiency of modern production.
2. Image Detection Applications
Inspection is one of the most important applications of machine vision in the industrial field. Almost all products require inspection, but manual inspection has many drawbacks. Manual inspection has low accuracy, which becomes even more difficult to guarantee over long periods of work, and its slow speed can easily affect the efficiency of the entire production process. Therefore, machine vision is also widely used in image inspection, such as: detecting characters on the edge of coins. In the fifth set of RMB banknotes issued in October 2000, the one-yuan coin had enhanced anti-counterfeiting features on its side. Due to the strict control requirements of the production process, a visual inspection system was installed in the final stage of coin minting; color registration and color matching checks in the printing process; printing quality checks on beverage bottle caps in the packaging process; barcode and character recognition on product packaging; and defect detection in glass bottles. Machine vision systems for defect detection in glass bottles also include pharmaceutical glass bottles, meaning that machine vision has also entered the pharmaceutical field. Its main inspections include dimensional inspection, bottle body appearance defect detection, bottle shoulder defect detection, and bottle mouth inspection.
3. Visual positioning applications
Visual positioning requires machine vision systems to quickly and accurately locate and confirm the position of the part being measured. In the semiconductor packaging field, equipment needs to adjust the pick-up head based on the chip position information obtained from machine vision to accurately pick up and bind the chip. This is the most basic application of visual positioning in the machine vision industry.
4. Applications in object measurement
The most significant characteristic of machine vision industrial applications is its non-contact measurement technology. It offers the same high precision and speed, but without the wear and tear of contact measurement, eliminating the potential for secondary damage caused by physical contact. Common applications include gear, connector, automotive parts, IC component pins, twist drills, and thread inspection.
5. Object sorting applications
In practice, object sorting applications follow the identification and detection process. Machine vision systems process images to achieve sorting. In industrial applications, machine vision is commonly used for food sorting, automated sorting of parts with surface defects, and cotton fiber sorting.
gongkong prediction
Previously, thanks to the booming industrial robot market, the machine vision market in China's industrial sector experienced a significant increase in 2010-2011, with a growth rate of around 30%.
According to Gongkong market research, the domestic industrial machine vision market has reached 2.6 billion yuan since last year, achieving significant growth. This is a result of the initial formation of my country's intelligent manufacturing industry chain. The electronics manufacturing industry is the main driver of this market growth. In the future, with the further advancement of intelligent manufacturing in China, the industry's demand for machine vision will gradually increase.
Gongkong predicts that the machine vision market will maintain a good growth trend, with an estimated growth rate of around 20% this year and a market size exceeding 3 billion yuan.
