The model training is done in Google Colab The model is then downloaded and perform object detection locally with the FYP.py
There are 3 parts in the FYP.py,
- Detect Objects
- Crop the object detected and classify them into their specific classes
- Relocated the bounding box of detected objects into the full image to determine if there is any false call or escapee