Intelligent product quality inspection, use AI vision technology to automatically identify the appearance defects of industrial products, and provide a software hardware integrated solution
Cooperation consulting
Scheme introduction
Scheme architecture
Application scenarios
Characteristic advantages
Related recommendations
Scheme introduction
Scheme introduction
Provide a complete set of industrial product appearance defect detection solutions by using AI in-depth learning core technology, combined with 2D/3D industrial cameras, automatic chemical equipment, mechanical arms and other hardware equipment
Application effect
It can be integrated with the existing production line of the enterprise to realize automatic quality inspection of the final finished products, intermediate process products, raw materials and other links, replace the existing manual quality inspection operations, open up the automatic process of the whole production line, and improve quality inspection and production efficiency
Advantages compared with traditional technology
Compared with the traditional AOI machine vision automatic detection technology, the artificial intelligence deep learning technology can detect more defect types, with higher accuracy, robustness and generalization ability. When new defect types are found in the production line, the model update can be faster and cheaper
Scheme architecture
Scheme composition and use process
Enterprise users select product types related to their own enterprise products in the application scenario, and launch cooperation applications. Baidu will implement solutions based on enterprise needs.The implementation process of the scheme includes model verification, supplementary training, integration implementation, continuous optimization and other stages
Model validation
The enterprise provides sample pictures of existing defects of products to verify whether the recognition effect of existing models meets the requirements
Supplementary training
If the recognition effect of the existing model is poor, the enterprise provides supplementary training pictures to optimize the model
Integrated landing
Investigate the production line of the enterprise on the spot, and formulate hardware supporting scheme for the production line. You can choose all-in-one machine or hardware customization
Continuous optimization
The new defect types in operation can be easily updated through online services
Application scenarios
Keyboard appearance inspection
Appearance inspection of wood floor
Inspection of industrial fuel injection nozzle
Textile needle inspection
Plastic bottle detection
Inspection of medical instructions
Glass bottle size inspection
Keyboard appearance inspection
In the keyboard manufacturing and assembly line, automatically identify the qualification of the keyboard after assembly, including missing parts, wrong assembly, and normal
Cooperation cases
Characteristic advantages
Software and hardware integration
Provide a full set of integrated landing scheme integrating software and hardware, which is convenient and fast
Leading algorithm model
Have the advantage of leading algorithm model, and create an optimal identification model for each subdivision
Edge accelerator card
Baidu has Edgeboard artificial intelligence edge computing accelerator card, which can realize on-site acceleration of AI models on end devices and realize high-speed real-time recognition of production lines
Online model training management
SaaS service for model supplementary training is provided to facilitate the supplementary training of new defects and model update after the quality inspection scheme is put into operation, so as to achieve high speed and efficiency
Welcome cooperation between enterprises and service providers
If the above scenario cannot meet your needs, or you have a good quality inspection cooperation scenario, please contact us quickly, and click the cooperation consultation below
