Eyes for Your Robot

The rc_visard sensor family enables robots to generate and process time and location-related data in real time.

This plug-and-play machine vision solution is designed to support robotic applications ranging from bin-picking to navigation. With its onboard processing capabilities it can be integrated directly, with typically no need for external computers.

rc_visard Stereo Sensor

The four versions of the rc_visard feature two different baselines (65 mm, 160 mm) and a color or monochrome acquisition capacity.

rc_visard 65 monochrome

Particularly suitable for robot/ arm-mounting, e.g. for pick-and-place tasks, in challenging lighting conditions.

rc_visard 65 color

Particularly suitable for robot/ arm-mounting, e.g. for pick-and-place tasks, when color differentiation is required.

rc_visard 160 monochrome
(in two variants)

Particularly suitable for static mounting and challenging lighting conditions; also available with a focal length (lenses) of 6 mm for set-ups with a larger viewing distance.

rc_visard 160 color

Particularly suitable for static mounting, e.g. in a robot cell, or mobile applications when color differentiation is required.

NEW! rc_visard NG: The flexible vision platform rc_visard NG allows the easy integration of proven software solutions with application-specific additions. The ‘next generation’ of the rc_visard is powered by today’s most advanced embedded AI solution, the NVIDIA® Jetson Orin™ series.

A Plug-and-Play Machine Vision Solution

rc_visard users do not need experience or expertise in machine vision:

Thanks to its user-friendly web interface, the rc_visard can be used intuitively and the installation effort is reduced to a minimum.

The rc_visard can also be connected and configured via Rest-API, ROS, and GenICam.

rc_visard accessories

The rc_randomdot is a specifically tailored projector that can be used as an optional enhancement to the rc_visard. It is recommended when the perception of particularly difficult scenes with little or no natural texture is required.

The rc_cube S is a the ideal edge computer for your rc_visard: It increases the performance of the rc_reason modules and is particularly useful for applications in which high-speed processing is of the essence.

A ConnectivityKit and CalibrationGrid are optionally available with your rc_visard. The modules of the rc_reason software suite optimize the sensor’s performance for specific applications.

rc_visard Features and Specifications




Camera Image

The image of the left camera serves as reference image for the image data set.

Depth Image

Distance from sensor to environment is computed through triangulation of the stereo image using the SGM methodology. It serves as input for point cloud generation and 3D reconstruction.

Confidence Image

The confidence image shows the confidence in each measured value ​​for further processing steps.

3D Reconstruction

3D reconstruction, calculated from the depth image, no texture.

Depth Image and Point Cloud

With an integrated graphics card, depth images are processed directly onboard the sensor and do not require any external computing power. This is an ideal prerequisite for mobile robotic systems. Point clouds can be easily generated from the depth images.

Position and Orientation

Using the most advanced visual inertial odometry algorithms, rc_visard records its current position and alignment with highest precision. The low latency allows for direct integration into the control loop when rc_visard is mounted on a robot.

Watch the Video

Indoor and Outdoor Use

rc_visard can be used reliably in indoor and outdoor areas.

Global Shutter

Accurate image data even with fast movements


Several sensors can operate  in a limited space without interference.

Technical Specifications

rc_visard 65 rc_visard 160
Base distance 65 mm 160 mm
Focal length (lenses) 4 mm 4 mm 6 mm
Image resolution 1280 x 960 Pixel (1.2 MPixel)
Field of view horizontal 61° 43°
vertical 48° 33°
Depth measurement range* 0,2 m – 1,0 m 0,5 m – 3,0 m
Workspace 0,2 m distance 0,17 m x 0,18 m
0,5 m distance 0,54 m x 0,45 m 0,44 m x 0,45 m 0,24 m x 0,30 m
1,0 m distance 1,14 m x 0,90 m 1,04 m x 0,90 m 0,64 m x 0,60 m
2,0 m distance 2,24 m x 1,80 m 1,44 m x 1,20 m
3,0 m distance 3,44 m x 2,70 m 2,24 m x 1,80 m
Depth resolution @ 0,2 m 0,04 mm
@ 0,5 m 0,2 mm 0,1 mm 0,06 mm
@ 1,0 m 0,9 mm 0,4 mm 0,3 mm
@ 2,0 m 1,5 mm 1,0 mm
@ 3,0 m 3,3 mm 2,2 mm
Depth image resolution & FPS 1280×960 Pixel (F) @ 1 Hz
640×480 Pixel (H) @ 3 Hz
320×240 Pixel (M) @ 15 Hz
214×160 Pixel (L) @ 25 Hz
Egomotion (IMU) 200 Hz, low latency
SLAM Onboard (optional)
GPU/CPU Nvidia Tegra K1
Power supply 18-30 V
Protection class IP54
Cooling Passiv
Temperature range (Celsius) 0°C – 50°C
Size (LxHxW in mm) 135 x 75 x 96 mm 230 x 75 x 84 mm
Weight 680 g 850 g

*In principle, the rc_visard can measure infinite ranges.  However, the accuracy is decreasing quadratically due to the stereo method, which is why we recommend the measurement range stated above.


The complete product documentation of the rc_visard is available online.


Roboception offers tools for easy and efficient interfacing and operation of the rc_visard.

Get in touch to learn more:

    Please prove you are human by selecting the Tree.


    Visit us.

    Roboception GmbH
    Kaflerstr. 2
    81241 Munich

    +49 89 889 50 79-0

    Send us a message.


    Hannover Messe 2017

    Vom 24 – 28. April 2017 stellen wir
    zu ersten Mal rc_visard vor.

    Besuchen Sie uns am Stand G23
    in der Halle 017.

    → Eintrittsticket gratis bekommen