Will Rtabmap act with Realsense r200 as well in reality as in Gazebo?

[Dan010123]

Hello, just want to clarify before bying the camera. The max depth range of the camera is about 15 meters (with respective error). I have a Rtabmap SLAM simulation in Gazebo with this camera. Is it close to the real functionality of Rtabmap with the real Realsense camera? Or should I limit "far" clipping to 15m to simulate it correctly? The Gazebo sensor config is following:

 <sensor name="camera" type="depth">          
        <always_on>true</always_on>
        <visualize>true</visualize>
        <update_rate>30</update_rate>
        <camera name="intel_realsense_r200">
          <horizontal_fov>1.02974</horizontal_fov>
          <image>
            <width>640</width> 
            <height>480</height>
            <format>R8G8B8</format>
          </image>
          <clip>
            <near>0.02</near>
            <far>300</far>
          </clip>
          <noise>
            <type>gaussian</type>
            <mean>0.0</mean>
            <stddev>0.007</stddev>
          </noise>
        </camera>
          <plugin name="camera_driver" filename="libgazebo_ros_camera.so">
            <ros> /ros>
            <hack_baseline>0.07</hack_baseline>
          </plugin>
      </sensor>

Thanks!

[matlabbe]

In Gazebo, you have perfect depth estimation, which is different than a real stereo camera (which is exponentially more noisier the farther from the camera). A lot of variables are different from Simulation versus Reality: the actual environment, lighting, textures, calibration may all be different, so it is hard to tell if it will behave the same.

Limiting to 15 meters may not simulate exactly like a stereo camera, at least if you are using rtabmap in stereo mode (recommended) so that "infinite" (very small disparity) depth features are still used for motion estimation.