Using positive MinDisparity in RTabMap standalone application

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|

Using positive MinDisparity in RTabMap standalone application

HiddenNetwork
Hi there,

I have been working with the RTabMap standalone application.

I tuned my StereoSGBM parameters outside of RTabMap GUI (using OpenCV). Due to the range of depth sensing I would like, I have parameters with a non-zero minDisparity.  The disparity map looks fine in OpenCV, however when using the same set of parameters in RTabMap, the disparity map and instantaneous pointcloud look incorrect.

For example, with these default settings on the demo video, the results look fine.

However, when changing to a positive minDisparity, it seems that for disparities < minDisparity, RTabMap's disparity map shows up as blue, and the instantaneous pointcloud shows the blue sections as a flat depth, which ultimately results in an incorrect map.

How can I use a positive minDisparity in the RTabMap standalone application?
Reply | Threaded
Open this post in threaded view
|

Re: Using positive MinDisparity in RTabMap standalone application

matlabbe
Administrator
Hi,

Thx for detailed report of your results. I fixed the output disparity when minDisparity is used.

cheers,
Mathieu