With the increasing requirements for product quality in the food/agriculture industry, packaging inspection is crucial. WeLinkirt's 2D AI AOI equipment brings a new solution to food packaging inspection with its unique technology.
User scenario: A leading food manufacturer's packaging process on the production line involves food packaging sealing, label pasting, and product inkjet coding. The detection objects are various food packages, including plastic and carton packages, which need to ensure tight packaging seals, correct label pasting, and clear and complete inkjet codes.
Pain points: Traditional detection methods have many problems in the detection of packaging seals, labels, and inkjet codes. The missed detection rate is about 3%, which causes some defective products to enter the market and affects the brand reputation. The false alarm rate is as high as 25%, increasing a large amount of re-inspection work. The labor cost is high, and a large number of manual inspections are required. In addition, with the rise of non - Transformer architecture technology, traditional methods are difficult to meet the needs of more efficient and accurate detection.
Technical Principle
WeLinkirt's 2D AI AOI equipment uses high-resolution 2D imaging technology and deep-learning secondary image judgment algorithm. High - resolution 2D imaging can capture micron-level details on the packaging surface, providing clear and accurate image data for subsequent analysis. The deep-learning secondary image judgment algorithm is based on a non - Transformer architecture. Through learning from a large number of positive samples and a small number of samples, it can quickly and accurately identify various defects on the packaging surface. The non - Transformer architecture has the advantages of high computing efficiency and fast model training speed when processing such planar image data. It can analyze a large number of images in a short time and has relatively low requirements for hardware resources.
- High - resolution 2D imaging: Through a high-precision industrial camera and optical system, a micron-level image resolution is achieved to ensure that small defects on the packaging surface can be clearly captured.
- Deep - learning secondary image judgment: The first image judgment quickly screens out the areas that may have defects, and the second image judgment uses semantic understanding and feature learning to make a more accurate judgment on these areas, effectively filtering false alarms.
- Advantages of non - Transformer architecture: It avoids the complexity of the Transformer architecture when processing simple planar images, reducing the training time and computing resource consumption.
WeLinkirt's Solution and Product
Centered around the 2D AI AOI equipment and combined with the DaoAI AI AOI software system. The 2D AI AOI equipment has the ability of high-speed online full inspection and can perform real-time detection on the food packages on the conveyor belt with a detection accuracy reaching the micron level. The DaoAI AI AOI software system uses APDT positive-sample/few-sample learning technology. With only 1-20 good samples, it can complete 0-code automatic programming in 5 minutes, greatly shortening the change-over time. At the same time, the semantic false-alarm filtering function of the system effectively reduces the false-alarm rate.
The 2D AI AOI equipment provides reliable guarantee for food packaging inspection with its high-speed and high-precision detection ability.
Quantitative results: After using WeLinkirt's solution, the detection rate increased to 99.4%, the missed detection rate decreased to <0.6%, effectively preventing defective products from entering the market. The false-alarm rate decreased by -63%, significantly reducing the re-inspection workload. The change-over time was shortened from several hours to 5min, improving the flexibility and production efficiency of the production line.
FAQ
What types of packaging defects can the 2D AI AOI equipment detect?
This equipment can detect planar defects such as loose packaging seals, incorrect label pasting, unclear or missing inkjet codes. It achieves micron-level accurate detection through high-resolution 2D imaging and deep-learning algorithms.
Is the change-over operation complex after using this equipment?
No, it isn't. Combined with the DaoAI AI AOI software system and using positive-sample/few-sample learning technology, only 1-20 good samples are needed. 0-code automatic programming can be completed in 5min for quick change-over.
How much can the false-alarm rate of the equipment be reduced?
In practical applications, the false-alarm rate can be reduced by -63%. This benefits from the deep-learning secondary image judgment and semantic false-alarm filtering function, effectively reducing unnecessary re-inspection workload.