In the current food industry, leveraging new visual perception solutions and AI-AOI technology to achieve more efficient quality control has become a hot trend. WeLinkirt's AI AOI software system demonstrates significant advantages in the nut/french fry color sorting process.
User Scenario: On the nut/french fry production line of a leading food manufacturer, the color sorting process is a crucial step to ensure product quality. The inspection objects are various nuts (such as cashews, almonds, etc.) and french fries, and products with non-standard color and shape need to be screened out to ensure the high-quality of the final products on the market.
Pain Points: In the traditional nut/french fry color sorting process, there are many quantitative dilemmas. On the one hand, the manual inspection has a relatively high miss-detection rate of about 5%, making it difficult to ensure product consistency and stability. On the other hand, the false-alarm rate is as high as 15%, resulting in a large number of qualified products being wrongly removed and increasing production costs. At the same time, manual inspection is inefficient, with high labor costs and a long change-over time of about 30 minutes, which cannot meet the needs of large-scale production and diverse products. In addition, with the increasingly strict food safety regulations, traditional detection methods are difficult to meet compliance requirements. Combining with today's hot-spot direction, traditional visual inspection solutions are difficult to achieve efficient quality control in the new production environment.
Technical Principle
WeLinkirt's AI AOI software system is based on an advanced visual foundation model with powerful feature recognition capabilities. It uses deep-learning algorithms to accurately identify the color and shape features of nuts and french fries. The system is trained with a large number of positive samples. Through the APDT positive-sample/few-sample learning technology, it can quickly and accurately learn the features of good products with only 1-20 good-product samples. During the inspection process, the system analyzes the images in real-time and uses semantic false-alarm filtering technology to mark products that do not meet the features of good products. At the same time, the system supports SDK/API/Docker deployment to achieve 100% local privatization, ensuring data security and production autonomy. This technical principle is effective because deep-learning algorithms can automatically extract complex features from images, the positive-sample/few-sample learning technology reduces the dependence on a large number of samples and improves training efficiency, and the semantic false-alarm filtering technology reduces the false-alarm rate and ensures the accuracy of detection.
- Visual foundation model: Conduct in - depth recognition and learning of various features.
- APDT positive-sample/few-sample learning: Quickly learn the features of good products and reduce sample requirements.
- Semantic false-alarm filtering: Accurately distinguish between good and bad products and reduce false alarms.
- Local privatization deployment: Ensure data security and meet the enterprise's independent production needs.
WeLinkirt's Solution and Product
WeLinkirt's AI AOI software system is the core product to solve the nut/french fry color sorting problem. It has a zero-code automatic programming function. Only one good product is needed, and programming can be completed in 5 minutes, greatly shortening the change-over time. The semantic false-alarm filtering function of the system effectively reduces the false-alarm rate and improves inspection efficiency. At the same time, the system supports SDK/API/Docker deployment and can achieve 100% local privatization, with data staying within the factory, ensuring the enterprise's data security. In addition, WeLinkirt's DaoAI 2D / 3D AI AOI equipment can be used as a supporting product. Its self-developed 3D camera and three-dimensional morphology reconstruction technology can detect hidden defects and micron-level morphology of products, further improving the accuracy and comprehensiveness of detection.
WeLinkirt's AI AOI software system provides reliable guarantee for quality control in the food industry with its efficient and accurate detection capabilities.
Quantitative Results: After using WeLinkirt's AI AOI software system, the nut/french fry color sorting process of the food manufacturer has achieved remarkable results. The detection rate has increased from the original 95% to 98.5%, greatly improving product quality stability. The false-alarm rate has been reduced by -60%, reducing the wrong removal of qualified products and lowering production costs. The change-over time has been shortened from the original 30 minutes to 5 minutes, improving production efficiency and meeting the production needs of diverse products.
FAQ
What is the accuracy of the AI AOI software system in inspecting nuts/french fries?
Based on the visual foundation model and deep-learning algorithms, the system has a detection rate of 98.5%. It can accurately identify color and shape features. Combined with supporting equipment, it can also detect hidden defects and micron-level morphology, ensuring inspection accuracy.
Is the system's change-over operation complex? How long does it take?
No, it isn't. The system has a zero-code automatic programming function. Only one good product is needed, and programming can be completed in 5 minutes, greatly shortening the change-over time and meeting the needs of diverse production.
How does the system ensure data security?
The system supports SDK/API/Docker deployment and can achieve 100% local privatization. Data stays within the factory, effectively ensuring the enterprise's data security.