What is machine vision?
Machine vision is a rapidly developing branch of artificial intelligence. Simply put, machine vision uses machines to replace human eyes for measurement and judgment. A machine vision system uses machine vision products (i.e., image acquisition devices, which can be CMOS or CCD) to convert the captured target into image signals, which are then transmitted to a dedicated image processing system. This system obtains the target's shape information and, based on pixel distribution, brightness, color, and other information, converts it into digital signals. The image system performs various calculations on these signals to extract the target's features, and then controls the on-site equipment based on the judgment results.
Classification of machine vision systems
• Smart camera
•Based on embedded systems
• PC-based
Composition of machine vision system
• Image acquisition: light source, lens, camera, acquisition card, mechanical platform
• Image processing and analysis: Industrial control host, image processing and analysis software, graphical user interface.
• Judgment execution: Telephone unit, Mechanical unit
How machine vision works
Machine vision inspection systems use CCD cameras to convert the targets to be inspected into image signals, which are then transmitted to a dedicated image processing system. Based on pixel distribution and information such as brightness and color, the signals are converted into digital signals. The image processing system performs various calculations on these signals to extract the features of the targets, such as area, quantity, position, and length. Based on preset tolerances and other conditions, the system outputs results, including size, angle, number, pass/fail, presence/absence, etc., to achieve automatic recognition.
Applications of machine vision
Case 1: Application of Robot + Vision for Automated Loading, Unloading and Positioning
There are two vibrating disks on site. Vibrating disk 1 moves the doll to vibrating disk 2, and vibrating disk 2 moves the doll from the reverse side to the front. This application uses the VD200 vision positioning system from Shenzhen VisionDragon. This system determines the front or back of the doll and sends the coordinates of the doll's front position to the robot via serial port. After receiving the coordinates, the robot moves to grasp the product. When many dolls in the vibrating disk are reversed, the VD200 vision positioning system needs to determine the number of reversed dolls. When there are too many reversed dolls, the VD200 vision system sends a command to vibrating disk 2 to move the reversed dolls to the front.
This positioning system determines whether the doll is facing forward by using small holes on its surface, calculates the coordinates of the doll's center point, and sends the results to the robot. The VD200 vision positioning system enables automated feeding, significantly reducing labor costs and greatly improving production efficiency.
Case Study 2: Application of Visual Inspection in Electronic Components
This product is a button component for electronic products. It arrives as a strip, with two rows of components on the strip. After locating each component, a speckle tool is used to detect the grayscale value of a fixed area on the product to determine if there is any missing adhesive in that area.
This application utilizes the DragonVision vision system solution from Shenzhen Vision Dragon Technology Co., Ltd., employing two cameras and a light source in conjunction with mechanical equipment to inspect eight double-sided products at a time, approximately 1500 products per minute. When a defective product is detected, an alarm immediately sounds and the machine shuts down, ensuring a high product pass rate and the normal operation of the equipment, thereby improving production efficiency.
Case Study 3: Brewery's Filling Level Detection System
As each beer bottle moves past the detection sensor, the sensor triggers the vision system to emit a strobe light and capture an image of the bottle. After acquiring the image and saving it to memory, the vision software processes or analyzes the image and issues a pass/fail response based on the actual fill level of the bottle. If the vision system detects that a beer bottle is not fully filled (i.e., it fails the detection), it will signal a deflector to remove the bottle from the production line. The operator can view the rejected bottles and ongoing process statistics on the display screen.
Current Status of Machine Vision Applications
Abroad, the widespread application of machine vision is primarily seen in the semiconductor and electronics industries, with approximately 40%-50% concentrated in the semiconductor sector. Specific applications include: PCB (Printed Circuit Board): various PCB assembly technologies and equipment; single-sided, double-sided, and multi-layer circuit boards, copper-clad laminates and related materials and accessories; auxiliary facilities, consumables, inks, chemicals, and accessories; electronic packaging technology and equipment; screen printing equipment and related materials. SMT (Surface Mount Technology): SMT processes and equipment, soldering equipment, testing instruments, rework equipment and various auxiliary tools and accessories, SMT materials, mounting plates, adhesives, fluxes, solder and anti-oxidation oils, solder paste, cleaning agents, etc.; reflow soldering machines, wave soldering machines, and automated production line equipment. Electronic manufacturing equipment: electronic component manufacturing equipment, semiconductor and integrated circuit manufacturing equipment, component forming equipment, and electronic molds. Machine vision systems have also been widely applied in various aspects of quality inspection, and their products occupy a crucial position in these applications. In addition, machine vision is used in various other fields.
In China, the application of vision technology began in the 1990s. Because the industry itself was relatively new, coupled with insufficient widespread adoption of machine vision products and technology, its application in the aforementioned industries was virtually nonexistent. Currently, most machine vision products in China are from foreign brands. Most domestic machine vision companies started by acting as agents for various foreign machine vision brands. As the application of machine vision has continued, these companies have gradually grown in size and their technology has matured.
With the improvement of economic levels, 3D machine vision has begun to enter people's field of vision. 3D machine vision is mostly used for grading fruits and vegetables, timber, cosmetics, baked goods, electronic components, and pharmaceutical products. It can improve the production capacity of qualified products and discard inferior products early in the production process, thereby reducing waste and saving costs. This function is ideally suited for imaging product attributes such as height, shape, quantity, and even color.
In terms of industry applications, machine vision technology is mainly used in pharmaceuticals, packaging, electronics, automobile manufacturing, semiconductors, textiles, tobacco, transportation, and logistics. Replacing manual labor with machine vision technology can improve production efficiency and product quality. For example, in the logistics industry, machine vision technology can be used for sorting and classifying express packages, eliminating the need for manual sorting that is common in many express companies. This reduces the damage rate of items, improves sorting efficiency, and reduces manual labor.
