Object_Detection_Deployed_on_Docker

Train and predict an object detection model on docker using tensorflow object detection 2.0

Steps to follow:

1. To build image, update the root directory of tensorflow and the name of the image in the Build.bat file.
2. Copy the Dockerfile from this repository to .\research\object_detection\dockerfiles\tf2\Dockerfile location.
3. Run the Build.bat file.
4. Clone this repo and create a folder name images and paste there all the training images along with annotations (xml) files.
5. Update the path of this repository directory in Run.bat file and confirm the ports as per the availability.
6. Run the Run.bat file to run docker image.
7. Follow the Final_OD_pothole.ipynb file to train and predict the object detection model.