I wanted some support RE navigating object detection for a mono camera. Object detection + bounding boxes work well for our Zed2i stereocamera, but I was wondering how to set it up for the mono camera. In the stereocamera yaml config files, there’s an object detection option (toggled between true and false) that I was using, but I don’t see this for common_mono.yaml file. How can I toggle it? I tried using the same block of code in the yaml file from the stereo file, but this doesn’t seem to work (ros2 topic list doesn’t show object detection).
I tried to look through your docs and don’t see any info on this - any support would be great. The mono cameras we are using are ZedXOne GS.
Thank you. What’s the rationale behind not providing object detection support out of the box for the mono cameras? I’m wondering if I can set up my own object detection (using something like yolo) or if this might not be possible at all with the cameras.
It’s not planned for the moment because it wouldn’t add any value with respect to using external libraries (i.e, YOLO).
The sl::Camera (with stereo processing) adds important 3D information to detection and tracking, but sl::CameraOne, at the current status, cannot add more than “simple” 2D detection.