seeed studio camera Vision AI Platform
Specifications
- Product Name: Camera 2002
- Processor: C906@1GHz + C906@700MHz
- System: RISC-V SoC SG2002
- AL Performance: 1TOPS@Int8
- CPU: 8051@8KB SRAM
- Memory: 256MB
- Video Encode: 5MP @ 30Fps
- Camera Sensor: OV5647 (version1)
- Storage: 8GB (version2) eMMC, expandable up to 64GB with removable TF card (not included)
- I/O: Type-C(2.0), Ethernet, Wireless, USB
- Audio: On-board Mic, External Speaker
- Dimensions: 40x40x36.5mm
The Vision AI Platform for Everywhere
Camara is an open-source and tiny AI camera, programmable and customizable, powered by RISC-V SoC, delivering on-device 1 TOPS AI performance with video encoding 5MP @30 FPS. The modular hardware design and expandable
interfaces offer the most versatile platform for developers building vision AI systems.
Modular Design Overview
The camera system consists of different boards that can be customized based on your needs:
- Base Board: Includes Type-C, SD Card, UART, Ethernet, and expandable interfaces
- Core Board: Contains CPU, RAM, eMMC, and Wireless module
- Sensor Board: Compatible with various camera sensors like OV5647, IMX335, and more
Features
YOLO Native
With Wireless and Different RAM Options
Modular Design
Smooth Software Interaction
Buildroot System for Programming
A lightweight customized Linux system: Support Python and Node.js directly from the console Deploy the compiled executive file from C or Rust. Multi-threading is supported, running all tasks without worrying about the conflicts.
Node-RED Integration for No-code Workflow
Customized nodes for the camera allow you to call the camera API and use the TPU to load models directly onto the device.
Simple steps as
Select node module Configurate node Deploy
Software Interaction
The system runs on a Buildroot Linux system and supports Python and Node.js for programming.
Follow these steps for software interaction:
- Select the node module for the desired functionality.
- Configure the node settings accordingly.
- Deploy the changes to the camera device.
Node-RED Integration
Node-RED allows for a no-code workflow to interact with the camera. Customized nodes enable easy access to the camera API and TPU for loading models directly onto the device.
Deploy Camera Everywhere
Everything at seeedstudio.com/recamera
FAQs
Q: Can I expand the storage capacity of reCamera?
A: Yes, you can expand the storage capacity up to 64GB using a removable TF card (up to 2TB).
Q: What programming languages are supported?
A: The system supports Python and Node.js directly from the console for easy programming.
Q: How do I reset the device?
A: You can reset the device by pressing the Reset Button located on the camera.
Documents / Resources
 |
seeed studio reCamera Vision AI Platform [pdf] User Guide reCamera Vision AI Platform, Vision AI Platform, Platform |