An issue with the current program is that it converts in real time, which means that long videos take long times to process.
-
- As a user, I want the ability to edit and crop the video so I can target motion extraction to times that specifically interest me. (can extract multiple times on same snip)
AND/OR
-
- As a user, I want the ability to provide the program timestamps so that I can target extraction without having to crop or edit the original video. (faster but have to keep providing time stamps if want to process the same segment of video)
I believe that the second option will be better for speed, as if we can provide the timestamps directly to opencv we can skip manually cropping the video. It might be worth adding a cut/crop utility to the video as it can be easily done to the script without much hassle.
These should be done as function, we can add time stamps to the functions to specify time frames and just have them default to the full time if they are not populated.
An issue with the current program is that it converts in real time, which means that long videos take long times to process.
AND/OR
I believe that the second option will be better for speed, as if we can provide the timestamps directly to opencv we can skip manually cropping the video. It might be worth adding a cut/crop utility to the video as it can be easily done to the script without much hassle.
These should be done as function, we can add time stamps to the functions to specify time frames and just have them default to the full time if they are not populated.