This repo includes notebook to show how using pretrained YoloV3 model object detction can be done on videos. In the notebook two ways are shown:
- Extracted images from videos using opencv and used YoloV3 implemented in ArcGIS API for Python to detect objectf of interest like person, vehicle, etc.
- Used predict_video method to detect objects directly on videos.
- Path to sample video: https://1drv.ms/u/s!Ag2pQcEMqYk5nlebj7rk0O4g0ux0?e=TTvfki
- Path to sample predictions: https://1drv.ms/u/s!Ag2pQcEMqYk5nlb3-K9Um3K9xfcP?e=SIPGcR