Deciding the range of frame used for the 3D reconstruction

[DenDen047]

When args.end_frame is set -1, the start_frame and end_frame are automatically defined with the following process:

max_idx = filtered_points_2d_df['frame'].max() + 1
for i in range(max_idx):
    if num_marker(i) == len(target_markers):
        start_frame = i
        break
for i in range(max_idx, 0, -1):
    if num_marker(i) == len(target_markers):
        end_frame = i
        break

line 812 on all_optimizations.py.

[zicodasilva]

It might be better if we use the start_frame as the trigger for the automatic frame selection. That way we can use end_frame=-1 as an indication to use the entire video. So we have the options to:

• Manually set start_frame and end_frame to values > 0.
• Use the entire video segment by setting end_frame = -1.
• Automatically select the "best" frames based on the first time (and last time) the markers are all visible by setting start_frame = -1.

Just a thought.

[zicodasilva]

As I understand it, this will select a start frame where all the markers are visible (from any set of camera view points) and do the same from the other side for the end frame. I think this is a good way to select the frames that would produce the best reconstruction.