AOI Capacitor Automated Optical Inspection Systems for 3C Assembly line

product background:

Machine learning refers to the use of algorithms to enable computers to dig out information from data like humans. Deep learning, as a subset of machine learning, uses more parameters and more complex models than other learning methods. Makes the model's understanding of the data deeper and smarter. In computer vision, a pre-trained image classification network is usually used as the feature extraction front end of an object detection and segmentation network. Using these pre-trained networks as front-ends can ease the training of the entire model and often helps achieve higher performance in a shorter time. KEYE TECH integrates the application experience of the outer packaging inspection industry into the capacitor appearance inspection industry. And for the first time to introduce deep learning algorithms in the capacitor appearance inspection industry. The algorithm uses human mathematics knowledge and computer algorithms to build an overall structure, uses a high-performance computer for a large number of simulation training, and then combines a large amount of training data to adjust the parameters, and finally realizes the recognition of the appearance of the capacitor. The KEYE AI capacitor quality inspection machine is checked by the industry's top algorithm designers in sample collection, template establishment, actual measurement and debugging, after-sales service and other links, so that deep learning can be successfully applied to the capacitor quality inspection industry.

Inspection principle:

The materials are arranged by vibrating plate, Linear vibration and non-destructive conveying, and the capacitance is stably and orderly transferred to the specific inspection station, and then the high-pixel sample image captured by the industrial camera is transferred to the image processing software. After the system analyzes and processes, it judges good and defective products, and controls the corresponding pneumatic components to eliminate the defective products, and finally collects qualified products.

Inspection details:

The inspection machine adopts 14 sets of high-pixel industrial cameras and high-performance stroboscopic light sources to carry out omni-directional visual inspections for product appearance defects. The inspection speed is 400-600 pieces/min. The equipment can realize 7*24 hours of all-weather operation, and it can be online automatically. Eliminate substandard products.

The whole set of inspection equipment includes mechanical, visual electronic hardware and inspection system software. The entrance and exit of the equipment are reserved for docking positions to facilitate on-site operations.

1, inspection standard

2, system composition:

(1) Mechanical part:

The mechanical part is the core component of the system, which carries all the transmission devices and electronic hardware (industrial cameras, light sources, industrial computers, touch screens, electrical boxes, etc.). The conveyor is controlled by a servo motor, and the image capture is accurate. The equipment is equipped with two defective product inlets to separate serious defective products (such as drum bottom, oblique leather tip, etc.) and general defective products.

(2) Inspection software:

The software is the master control platform of the equipment, and its specific functions are as follows:

• Enabling and disabling the system detection function.
• Adjustment of the parameters and accuracy of detection indicators.
• Template sampling and storage, sample detection information storage and recall.
• Camera calibration
• Display system real-time production information (qualified rate, total output, speed, etc.). Browse, export and store historical production information.
• l Perform authority management according to the login account (for example, general operating users are prohibited from using parameter adjustment permissions)
• Operation and adjustment of the mechanical parts of the system (such as: air valve jet time adjustment, camera trigger time adjustment)
• Additional functions added due to customer needs

(3) Hardware part:

The hardware part includes cameras, lenses, light sources, high-speed air valves, industrial computers, touch screens and other electrical devices. For details, please refer to the equipment configuration details.

The overall logic control of the system is implemented by a high-speed FPGA chip.

Specific functions include:

• Trigger the light source strobe
• Trigger the camera to take a photo
• Trigger the solenoid valve to jet off waste products
• Control the belt speed
• Emergency treatment for system jams or similar situation
• Trigger the alarm light signalØ

3, Configuration:

4, Real-time image

Sample deformed

Bottom contamination

Sample deformed

Top bulged