rp-vio
rp-vio copied to clipboard
Run Real time with your own camera
Hi thanks for sharing your work. I want to run the RP-VIO in real time on pixhawk drone. How can provide mask(segmented) plane of camera? I am using mynt-eye and realsense 435i camera thanks
Hi,
You have multiple options. If you have depth information, the best would be to use a geometric plane-fitting method like plane_seg. If you want to use only RGB information, you could use a plane segmentation model like the PlaneRecover model provided by us. There could be newer and more efficient models for monocular plane segmentation, but I am not up to date. Lastly, if it's a scene with known planar regions like roads, you could also use a generic segmentation model like Faster-RCNN and just provide the plane class labels in the config file as shown in the VIODE example.
For passing these segmentations to RP-VIO, you need to publish the segmentation images on a topic, and specify it in the config file as mask_topic
.