Our program that allows you to select which object detection model to use to detect objects.
We currently support the models:
The SSD model is taken from here.
- Python >= 3.79
- OpenCV2
- PyTorch >= 1.6.0
- Torchvision >= 0.7.0
To use the video function place .mp4 videos into the folder media/DrivingClips.
To use the detector with for example SSD Mobilenet on file video.mp4, type:
run.py -ssdm video
To use it with the webcam just ommit the filename:
run.py -ssdm
Running the command
run.py
without any arguments just opens the webcam and displays its output.
To write information about the model like the minimum detection confidence or inference time to a file, use the optional argument -l. Example:
run.py -ssdm video -l