Python Motion Detection

One common issue with many long video recordings (such as from security or nature cameras) is processing the media into a usable format. This software provides a simple and intuitive user interface that takes long videos and outputs short subsections containing motion that is above a certain threshold.

The video is run through a scene change detection algorithm, and a temporary file is produced containing the resulting information. Then, each frame is given a median score based off of the detected motion. Once a certain threshold is met, a smart copy algorithm creates a dynamic start and end point, and those sections are copied from the video and saved into the output folder. The last step is repeated for all subsections that fit the motion criteria. Since the video section is copied, there is no re-encoding or quality loss.