 |
Product Introduction
Direct output of recognition results instead of just video streams ● Edge AI Inference: Zero load on the main controller CPU ● Structured Data Output: Objects / Coordinates / Events ● High Real-time Performance: Low latency processing |
Application Scenarios:
● I want to skip vision algorithm research and get direct results
● The main controller is overloaded; I need to offload vision processing
● For robot integration, I need a plug-and-play vision module
Main Specifications
| Depth Rating | 300m |
| Weight in Air | |
| Weight in Water | |
| Module Model | OAK-1-W |
| Image Sensor | IMX378 |
| FOV | 120°D/95°H/70°V |
| Resolution | 12MP (4032×3040) |
| Max Frame Rate | 60 FPS |
| Focal Length | 4.81mm |
| Aperture | 2.8 ±5% |
| Focus Range | FF: 60cm – ∞ |
| Lens Size | 1/2.3” |
| Pixel Size | 1.55µm x 1.55µm |
| Shutter Type | Rolling Shutter |
| Protocol | Ethernet |
| Input Voltage | 12VDC |
| Power Consumption | 2.5W~3W |
| Packing List | AI-M Underwater Camera *1, Standard 2m seawater-resistant cable (Length customizable up to 70m, contact customer service for longer distances) |
| Software Dev | ORS / Python / C++ |
Main Dimensions
Wiring Instructions
| Black | GND |
| Red | +12VDC |
| Blue | Reserve |
| Blue/White | Reserve |
| Green | Rx- |
| Green/White | Rx+ |
| Orange | Tx- |
| Orange/White | Tx+ |
Model Selection Guide
| Resolution | Zoom | Power | Communication | Target Recognition | Depth Perception | |
|---|---|---|---|---|---|---|
| RPI | 8MP | × | 12VDC / 15W | Ethernet | √
(Secondary Dev) |
× |
| GoPro | 27MP | × | 12VDC / 8W | Ethernet | × | × |
| AI-M | 1.55µm x 1.55µm | × | 12VDC / 3W | Ethernet | √ | × |
| AI-D | 1.55µm x 1.55µm | × | 12VDC / 3.5W | Ethernet | √ | √ |
| Zoom | Approx. 2.13MP | 20x Optical Zoom
12x Digital Zoom |
12VDC / 3.6W | BNC
(75Ω Coaxial) |
× | × |
| Observer | 4MP | 4x Optical Zoom
16x Digital Zoom |
12VDC /
PTZ Camera: 10W Camera + 4 Lights: 60W |
Ethernet | × | × |
