Configure yolo version 7 to save detections with video and image
Budget: $10 – $30 USD
Hello,
i want to configure the yolov7 in order to do this:
i provide to --source a video or live stream
They save:
in the folder img: one image per detected object
in the folder video: one small video of the detected object
in the folder txt: one txt (csv) file with information about the detections (for every detection they can have: timestamp, name od detected items, name of related video, name of the related image)
Important, the videos can have size 20 sec and this must be configurable via varian.
Also in case of detection:
Save in csv the details
Save in images the first frame of the detection
Save in video the next 20 sec after the detection
But for 20 second not add more events, so if at 13:50:20 detect a person, until 13:50:40 they ignore extra detections may have and keep only one csv entry/image/video.
i want to configure the yolov7 in order to do this:
i provide to --source a video or live stream
They save:
in the folder img: one image per detected object
in the folder video: one small video of the detected object
in the folder txt: one txt (csv) file with information about the detections (for every detection they can have: timestamp, name od detected items, name of related video, name of the related image)
Important, the videos can have size 20 sec and this must be configurable via varian.
Also in case of detection:
Save in csv the details
Save in images the first frame of the detection
Save in video the next 20 sec after the detection
But for 20 second not add more events, so if at 13:50:20 detect a person, until 13:50:40 they ignore extra detections may have and keep only one csv entry/image/video.
Related categories:
Python
Machine Learning (ML)
Artificial Intelligence
AI (Artificial Intelligence) HW/SW