Defining the number of image sequences in run_on_recorded_sequence

[mnl12]

I am trying to run ./run_on_recorded_sequence after the built. However, I get the error "Could not read image from ../../../M3T/data/_sequence/./color_camera_image_202.png", it seems that there is only two sequences of images exist. How can I define the number of sequences? Thanks.

[AleTarsi]

I have the same problem, did you manage to solve it?

• I’ve noticed that once you run ./gtest_run the software correctly generates detection pictures and stores them in the /temp folder.

• Instead, If I try to run:

./run_on_recorded_sequence ../../data/_sequence/color_camera.yaml ../../data/_body/triangle.yaml ../../data/_body/triangle_manual_detector.yaml ../../temp/

from the folder ~3DObjectTracking/M3T/build/examples I get exactly your error.

[CVtracker]

Having the same problem. Have you found a solution for this error?

Edit:
This is not an error it's a warning. You need more frames. In the example only 2 frames are provided, that's why the program exits with this warning.

If you create your own frame series let's say of 10sec of video, you can see your recording being played in the viewer. You can then start the tracking by pressing X to first start with manual detection, hit enter and you will see the tracking result as an overlay. After these 10sec are over, it again shows the same warning. Consider it like your camera just stopped the feed.

Hope that helps!