1TOPS Computing Power, 4-POE Port Development Board

When using POE cameras for human motion recognition, close-range identification, or AI-based commodity category recognition—such as deploying 2 to 4 POE cameras at upper and lower positionsin visual cabinets or unmanned vending cabinets for identification—you need a development board with sufficient computing power thatcan simultaneously support power supply for 4 POE cameras and Ethernet ports.

The editor at Dingchang believes that 1TOPS computing power is generally sufficient, supporting up to 4 cameras with around 480p or 720p resolution to work simultaneously. It can be used for developingcommodity recognition algorithms. Dingchang provides a basic API platform to help partners familiarize themselves with the entire development process of AIcommodity recognition, including automatic identification of food and beverages, fresh produce, packaged items, snacks, and other related commodities.

Dingchang's AI Basic Development APIPlatform offers complete methods and processes for multi-category commodity dataset creation, annotation, training, and on-board inference. New andexisting customers can quickly master the AI development process on the RK platform, using it for learning and research, or for secondary development forcommercial purposes.

Other features include 2 lock control interfaces, 1 human sensing interface, onboard 4G communication, support for GPS, ADC-IN, and Type-C USB.

This development board can be used todevelop devices such as Android industrial computers for visual storage cabinets, and can also interface with face recognition all-in-ones to support practical functions like QR code scanning and face payment.