E2 AOI Detection Principle, AOI Working Principle
E2 AOI Detection Principle, AOI Working Principle
With the popularization of SMT technology and the densification and miniaturization of SMT components, automatic optical inspection equipment (AOI) is being used by the majority of electronics manufacturers to monitor and ensure product quality.
Compared with manual visual inspection, AOI has higher repeatability and faster inspection speed. Studies in the 1980s showed that when two people checked the same board four times, their mutual identification rate was less than 28%, and only about 44% of them agreed with themselves. Even so, before 2005, most electronics manufacturers still relied on manual visual inspection.
Because of the early introduction of imported AOI equipment, friends in the electronics manufacturing industry feel that the use is cumbersome, complicated, and expensive; or because AOI equipment programming and debugging are complicated, engineers cannot give full play to the performance of AOI equipment, resulting in AOI failure to achieve The expected detection effect makes AOI equipment just a kind of "expensive furnishings." This article will discuss the working principle of AOI, how to evaluate the AOI system, and how to configure the AOI system according to the specific situation.
1. Understand AOI and its working principle
1. Definition: AOI-Automated OptICal Inspection is an automatic optical inspection device applied to the production line of SMT-Surface Mounted Technology, which can effectively detect printing quality, mounting quality and Solder joint quality. By using AOI as a tool to reduce defects, errors can be found and eliminated early in the assembly process to achieve good process control. Early detection of defects will avoid sending defective products to the later assembly stage. AOI will reduce repair costs and avoid scrapping unrepairable circuit boards.
2. Main features:
1) High-speed inspection system, not affected by PCB placement density; 2) Fast and convenient programming system, graphical interface, what you see is what you get, use placement data to automatically program inspection programs; 3) Combine for different inspection items Optical imaging processing technology has different detection methods (detection algorithms); 4) When the mounting position of the detected component is offset, the detection window will be automatically positioned to achieve high-precision detection; 5) Display the actual error image , It is convenient for workers to carry out the final visual check; 6) Statistic NG data analysis leads to bad reasons, and real-time feedback of process information.
2. The working principle of AOI
After more than ten years of development, the technical level of AOI is still in the stage of rapid development. How to achieve the best detection results has always been a technical topic for various AOI manufacturers. At present, there are many AOI brands visible in the domestic market, and each AOI has its own strengths; the AOI advantages of each brand mainly depend on its different innovative core software algorithms. The software algorithms usually used are: template comparison, edge check, Grayscale model, feature extraction, solid-state modeling, vector analysis, graph pairing and Fourier analysis, etc., although the algorithms are different, the operating principles of AOI are basically the same.
The tower-shaped lighting system provides 360-degree omnidirectional illumination of the inspected components, and then uses a high-definition CCD camera to collect images of the inspected components at a high speed and transmit them to the computer. The dedicated AOI software compares according to the inspection programs that have been compiled. Analyze; determine whether the tested components meet the pre-ordered process requirements. Simply put, the process of AOI inspection of components is to simulate the visual inspection of SMT components by workers, which is to automate, intelligentize and program manual visual inspection.
Image acquisition is to use a CCD camera to convert the light signal on the surface of the object into an electrical signal and send it to the image acquisition card. The image capture card digitally sends the image to the computer. This process is intuitive and easy to understand. After sending the image to the computer, how does AOI check the quality of the component? Let's take ALeader AOI as an example:
1. Optical principle:
The ring tower-shaped three-color LED light source is used for lighting, and the red (R), green (G), blue (B) and white (W) light obtained by the combination of three colors are projected on the PCB from different angles. The components are illuminated in 360 degrees. Different color information of the component body, solder joints, and pads are obtained through light reflection, bevel reflection, and diffuse reflection.
The surface of the pad is smooth, the red light is specularly reflected on the surface, and most of the blue light is reflected, so the color of the pad is red or yellow-white.
Both blue and yellow light are diffusely reflected on the surface. According to the principle of toning, blue and yellow are blended to obtain white light, which is equivalent to the original body receiving white light, so the original body is displayed in its own color.
The solder joint (solder paste) usually forms an inclined plane, so that most of the yellow light is reflected through the inclined plane, and the blue light enters the lens through reflection, so the color of the resulting solder joint is blue.
2. The principle of image comparison:
It is captured by a CCD camera, and then subjected to image processing (that is, it is converted into the digital signal we need according to the pixel distribution, brightness and color information). Pass these digital signals through a certain mathematical calculation method to obtain a standard error threshold, and then compare the threshold obtained from each tested image with the corrected standard threshold in the system. If the comparison result is less than the standard threshold Then the image passes the inspection, otherwise it is judged as unqualified.
3. Statistical modeling technology:
The above-mentioned image to be tested needs to be compared with the standard image in the system, then the standard image is obtained through "statistical modeling". Usually, ALeader AOI learns a series of qualified image templates, allows the computer to automatically memorize the general characteristics of all OK images, obtains the shape changes of the image and the characteristics of the changes that may occur in the future, and generates a diversified qualified image model.
3. How to evaluate and configure the AOI system:
We know that the current AOI is divided into offline AOI (Off line) and online AOI (In line). In fact, the specific use of online AOI or offline AOI must be weighed according to its own actual situation; if it is a small batch, multiple models, offline AOI is the best choice for manufacturers with frequent line changes, because the detection speed can meet 1.5 high-speed The patch line needs to be easily moved, and can flexibly respond to the inspection needs of any process; the online type is often fixed in a certain process inspection, and is generally used for long-term fixed variety detection, so as to save the time of program debugging. Improve the utilization rate and stability of the equipment.