Second-generation Deep Learning System
Camera board for Deep learning systems
Specification:Camera board dedicated to deep learning systems Function Overview: Captures image data of the material and, after processing, transmits the information to the deep learning computing board, the touchscreen, and the ejector driver board. Features: Equipped with a high-definition RGB sensor, the system captures and processes image data at high speed; combined with a deep learning processing unit, it accurately identifies and rejects materials—such as foreign
Constant-current source driver board
Specification:High-current material illumination control board with backlight driver and control signal module. Features: Each output channel supports a high maximum output current, allowing a single port to control the material indicator light for one channel of a 12-chute machine. The output current remains stable across the entire range, from low to high levels. The background light can be controlled by sending signals from the control module directly to the background
AI board
Specification:Image processing board integrating a chip with 2.5–10.0 TOPS of computing power; configurable to meet specific customer requirements. Features: It employs deep learning algorithms to perform multi-dimensional material recognition—covering attributes such as color, shape, texture, area, and brightness—thereby achieving high-precision sorting. Its hardware configuration supports capabilities such as real-time image processing and high-speed decision-making.
Switch board
Specification:5-Port Gigabit Ethernet Switch Features:Five 10/100/1000 Mbps auto-negotiation RJ45 ports Auto-MDI/MDIX; full-duplex and half-duplex modes MAC address auto-learning and aging; non-blocking, full wire-speed forwarding
HMI Board for Deep Learning System
Specification:For AI model Function:It processes touchscreen inputs and camera data, receives data transmitted by the machine, converts it into natural language, and outputs it to the touchscreen to facilitate communication with humans. Features:Equipped with a wireless module, it supports wireless communication, allowing users to remotely control the machine via devices such as smartphones and computers.