Accurate depth on reflective and dark/black surfaces.
High precision and repeatability across measurements.
Delivers detailed point clouds for enhanced 3D reconstruction.
Stereo depth, object detection, and segmentation run on the integrated NPU.
Enables fast object detection and segmentation at the edge.
Python/C++, OpenCV and ROS 2 packages for rapid development.
The NetraVis 1MP Global Shutter Stereo Depth Camera gives indoor robots and automated systems the precise 3D vision they need to accurately navigate their surroundings. Built with twin 1MP global shutter sensors on a 100 mm baseline alongside an integrated 6-axis IMU, it delivers depth data up to 6 meters. By keeping depth error under 1% @ 2m and wiping out motion blur, it provides steady, reliable spatial mapping and obstacle detection even in fast-moving operational environments.
It hooks up easily over USB 3.0 Type-C and plays nicely with Linux. Its compact, low-power design makes it ideal for AMRs, ASRS, collaborative robotic arms, and industrial tasks requiring precise Depth-to-Color (D2C) registration.
| Parameter | Value |
|---|---|
| 3D Technology | Active stereo |
| Depth sensors | Two monochrome, global shutter |
| Baseline | 100 mm |
| Depth accuracy | ≤1% @ 2 m |
| Depth range | 0.38m - 6m |
| Depth FOV (H × V) | 88° × 67° |
| Depth full resolution | 1080 × 720 |
| Depth frame rate | 50 fps @ 1080×720 |
| Depth output format | 16-bit Z-depth map (mm, unsigned short); disparity / point cloud / colorized depth via SDK |
| Depth compute | On-board ASIC |
| Wavelength | 850 nm IR filter |
| Parameter | Value |
|---|---|
| RGB sensor | Single, global shutter, 1 MP |
| RGB resolution | 1280 × 720 @ 30 fps |
| RGB FOV (H × V) | 97° × 67° |
| RGB output format | BGRa |
| Module output | Depth, IR and RGB |
| AI accelerator | 3.2 TOPS on-camera NPU |
| Supported model types | Object detection, segmentation, classification* & face detection* |
| Model support | YOLO26 |
| IMU | 6-axis |
| Interface | USB 3.0, Type-C |
| Focus | Fixed focus |
| Parameter | Value |
|---|---|
| Dimensions | 124 × 27 × 25.5 mm (±0.5 mm) |
| Weight | 109 g |
| Parameter | Value |
|---|---|
| Operating voltage | 1.2 A or 1.5 A @ 5 V |
| Typical power consumption | ≤4.5 W |
| Operating temperature | 0°C to 50°C |
* - Under Developement
Depth, IMU, RGB and point-cloud sample code using the our SDK, plus OpenCV demos.
Applications covering depth, IMU, RGB, OpenCV, point cloud and on-camera AI model inference.
Full ROS 2 packages
README files and setup/API documentation for fast bring-up.
≤1% deviation at 2 meters.
0.38m - 6m in full resolution mode.
Yes. A 3.2 TOPS on-camera NPU runs object detection, classification, segmentation and face-detection models -including YOLO26 -independently of the host processor.
USB 3.0 over a Type-C connector.
A Python wrapper, C++ sample applications, ROS 2 packages, and OpenCV demos, with accompanying documentation.
| Product Name | Platform | Wavelength | Suitable for | Interface | Action |
|---|---|---|---|---|---|
| NetraVis_USB | x86/ARM | 850 nm | Indoor | USB 3.0, Type-C |
|
| NetraVis_GMSL2 | NVIDIA AGX Orin / Thor | 940 nm | Outdoor | GMSL2 |
|
Mr. Thomas Yoon
+82-10-5380-0313
Employee Login
