What is an AOI Inspection System? What are its features?

AOI inspection system, also known as automatic optical inspection technology, also known as machine vision inspection technology or automatic visual inspection, is a device based on optical principles to detect common defects encountered in welding production. When it is detected automatically, the machine automatically scans the PCB through the camera. Collect images, compare the tested solder joints with the qualified parameters in the database, and after image processing, check out the defects on the PCB, and display/mark the defects through the display or automatic flags for maintenance personnel to trim.

The AOI detection system consists of a simple optical imaging and processing system integrated by general devices such as cameras, lenses, light sources, and computers. Under light source illumination, the camera is used to image directly, and then the detection is realized by computer processing. The advantages of this simple system are low cost, easy integration, relatively low technical threshold, and can replace manual inspection in the manufacturing process to meet the requirements of most occasions.

According to different imaging methods, AOI can be divided into three-dimensional (3D) AOI and two-dimensional (2D) AOI, three-dimensional AOI is mainly used for the measurement of object shape geometric parameters, part grouping, positioning, identification, robot guidance and other occasions; Two-dimensional AOI is mainly used for product appearance (color, defects, etc.) detection, classification of different objects or appearances, detection and classification of good and defective products, etc.

However, for the visual inspection of large-format or complex structural objects, due to the mutual constraints of field of view and resolution (or accuracy), or the special requirements of the production cycle for the detection speed, the AOI system composed of a single camera is sometimes incompetent, so it may be necessary to have multiple basic units integrated together to work together to complete the difficult inspection task. That is, an AOI system integration architecture with multi-sensor imaging and high-speed distributed processing is adopted.

The AOI detection system independently developed by Xinxiwang has the following characteristics:

1) Tool simplification

Users can change the tool modules in the visual framework according to the specific needs of the project, and the output parameters of the interface can also be set according to the actual needs. It is easy to use for the operator, which greatly shortens the on-site commissioning time of the project and improves the development efficiency of the project. At the same time, it is also conducive to the later maintenance of the project, and only a simple training of on-site operators is required to deal with the abnormalities in the actual production process.

2) Functional modularity

According to the processing process, the program is divided into several functional modules in the AOI system framework, and these functional modules only need to change the input and output parameters of the specific visual tools and modules in them to regenerate a new program. The modularization of the system framework function greatly simplifies the difficulty of the framework, which is more convenient for users to operate, and the divided modules are easy to be reused between different programs, which can reduce development time and improve development efficiency.

3) Intelligent processing

Under the development characteristics of miniaturization, high density, rapid assembly and variety diversification of the products to be tested, the amount of detection information is large and complex, and the requirements for detection accuracy and operation speed are getting higher and higher. The new Xiwang AOI inspection system not only meets the requirements of high-speed inspection in complex environments, but also has deep learning functions, which can tolerate deviations and unpredictable defects for the detection of surface defects, texture and material classification, assembly verification and deformation part positioning, character reading and distortion printing, etc., with reliable, easy to match and high-speed characteristics.

4) Data visualization

In the actual production process, pictures and data can be stored according to requirements. The images that can be stored include original pictures, screenshots, NG pictures and other pictures, and can simulate the actual production process and analyze product data according to the stored pictures. The storable data can be set in the system framework and can be stored according to the actual production. The software can record the data of the production process in real time, which is more in line with the requirements of Industry 4.0, and achieves real-time production data, accurate data, and key data focus. The software also has SPC (Statistics Process Control) function, which can draw a control chart for analysis for the actual production data, and judge the abnormality of the entire production process and the law of product data through the control chart. At the same time, the use of statistical data to monitor the production process is conducive to stabilizing production and improving yield.

5) Convenient operation

The user-friendly operation interface can set the number of analog cameras according to actual needs, and multiple cameras can be operated at the same time. After the delivery of the system software, only a simple training of the operator is needed to deal with the subsequent production line abnormalities and production alarms, and the on-site line watchers are not required to have professional visual knowledge, which greatly reduces the later system maintenance costs and processing time. In particular, when the production line is replaced with a new product, it only takes a small amount of time to replace the new product, and there is no need for a long downtime.