Dockerized Flask App to segment cracks and detect potholes on road images.
To create docker image of the project:
sudo docker build -t app .
To run the image created.
sudo docker run --name app -p 5000:5000 app
The containter is up and running on port 5000. Use the CURL command to send an image as a POST request to the Flask App running on localhost. The result is a JPEG image.
- For Cracks (U-Net CNN) :
curl -k -X POST -F 'image=@image_path/ -v http://0.0.0.0:5000/segment > crack_result.jpeg
- For Potholes (R-CNN) :
curl -k -X POST -F 'image=@image_path/ -v http://0.0.0.0:5000/detect/rcnn > pothole_result.jpeg