3DV-A Stereo System

This product has been discontinued.

3DV-A is a stereoscopic system able to provide full 3D reconstruction of the scene framed by two cameras.

Thanks to VisLab’s proprietary algorithms developed throughout the years, 3DV-A is able to provide a quick interpretation of the environment, detecting and tracking obstacles and estimating terrain surface, and slope..

3DV-A is available to power third parties robotics/vehicular/rover platforms and is sold together with samples of crossplatform code to ease its integration on existing systems.

3DV-A has the ability to provide a dense 3D map of the surroundings (a 3D point cloud of up to 3 millions 3D points per second) in real time. Version A (3DV-A) uses an unboxed board as its processing engine, to ease its installation even on board of small robotic platforms.

Technical Details

Hardware Details

Weight ~4 Kg
PC case Yes*
Interface Ethernet
Power supply 12 or 24 V DC
Power consumption Approx. 100 W
Warranty 1 year

*3DV-A is provided with case. A case-less solution can be purchased as well.


Function details

Sensor type Mono / Color (RGB)
Minimum operational distance 1 m
Minimum installation height 0.5 m
Outdoor applicable Yes, the camera enclosure can resist to dust or light rain
Output format UDP packets over Gigabit Ethernet connection
Output data rate 10 Hz
Output two raw images (left and right), raw 3D depth map (one distance estimation per image pixel)
Plugins Obstacle Detection (list of detected objects, tracked over time);
Terrain Mapping (as a heights grid);
Visual Odometry;
Pedestrian Classification;
Vehicle Classification;
Supported frameworks ROS, RTMaps

Configuration options

ID Resolution Depth Map Baseline [cm] Lens [mm] FOV Min Dist. [m] Accuracy [m] @ distance [m]
close medium far
CS1 1280×960 640×480 24 3.8 66° 1.8 0.01 @ 2.1 0.5 @ 15 1 @ 21
CS2 1280×960 640×480 12 3.8 66° 1 0.01 @ 1.5 0.5 @ 11 1 @ 15
CS3 1280×960 640×480 24 6 43° 3 0.01 @ 2.8 0.5 @ 20 1 @ 28
CS4 1280×960 640×480 12 6 43° 1.5 0.01 @ 2.0 0.5 @ 14 1 @ 20
CS5 1024×768 512×384 12 6 43° 1.3 0.01 @ 1.8 0.5 @ 12 1 @ 18

3DV-A output examples

When watching these videos, don’t forget to select HD quality in your YouTube settings.
Graphical rendering is obtained with VisLab’s framework GOLD.

Examples of 3D point cloud; the virtual camera moves into the point cloud to show the accuracy of the reconstruction.


Examples of 3D reconstruction, object detection and tracking, and terrain mapping.


On April 2, 2014, AutonomouStuff hosted VisLab’s webinar on “3DV: A Tool for World Reconstruction through Artificial Vision”. Click here to see it on VisLab YouTube Channel.

The 3DV embedded version (version 3DV-E), based on an FPGA, is now available. 3DV-E processes 640×480 pixels images at 30Hz and provides full size 3D point clouds in real time. Find more information here



Description Size
pdf3DV-A datasheet 500KB