OpenNCC is a production vision system with AI vision hardware and software for real-time ML application processing; tightly-integrated with a high-quality vision camera, ISP, video processing, and communications infrastructure.
Users can customize pre-trained AI/ML vision models and easily deploy their models on OpenNCC.
OpenNCC utilizes Intel's Myriad X Vision Processing Unit (VPU) for AI vision acceleration and supports the Intel Distribution of OpenVINO Toolkit that automatically maximizes ML model inference performance.
AI | |
---|---|
Performance | 4TFlops (Max) |
Model support | all of the OpenVINO models |
Framework support | ONNX, TensorFlow, Caffe, MXNet, Kaldi |
SOFTWARE | |
Image signal process | supported |
Open source resources | camera development pack OpenNCC CDK, technical documents, configuration tool OpenNCC View |
Language support | C / C++ / Python |
SDK functions | 1. retrieving video stream 2. AI models download and change 3. obtain model operation result 4. camera take photos, reset, etc. |
OpenView functions | configure camera setting, configure camera local AI model |
OpenNCC CDK OS support | Linux and windows |
OpenNCC View OS support | Linux |
HARDWARE | |
Dimensions | 50 × 50 × 55mm |
Weight | camera net weight 140g |
VPU | Intel Movidius Myriad X MV2085 |
Operating temperature | 0-50℃ |
RAM | 4GB |
Data port | USB Type-C 2.0/3.0 |
Power supply | 5V / 2A |
Camera module | 8MP visible light module |
Resolution | 3872 × 2180 |
Frame rate | 30Hz |
Horizontal FOV | 58° |
3872 × 2180 HD Resolution, With Image Pre-Processing Function
Video Output Formats Including YUV, H.264, H.265, MJPEG