The ROS output image is a circle not a normal rectangle image, And a viso problem

Issue #28 resolved
wu NL created an issue

Is that right? circle_mul.png

And when I use Visual Odometry as you said in https://bitbucket.org/crl/multisense_ros/issue/22/visual-odometry-using-multisense, the viso node warn "The tf from '/multisense/head_root' to '/multisense/left_camera_optical_frame' does not seem to be available, will assume it as identity!"

How could I fix it?

Thanks.

Comments (4)

  1. Matt Alvarado

    Hello,

    The image you attached is expected from 4 MP versions of MultiSense cameras. The large amount of vignetting is caused by the lenses which are intended to be used with smaller optical format imagers. A full explanation for this can be found at (https://support.carnegierobotics.com/hc/en-us/articles/203456457-2-megapixel-or-4-megapixel-)

    In terms of the viso2 TF errors can you please attach the frames.pdf output from running “rosrun tf view_frames”.

    Thanks, Matt Alvarado Engineer Carnegie Robotics

  2. Log in to comment