ROI Cavity Detection

This project aims to showcase cavity detection on Regions of Interest (ROI) from periapical dental radiographs. This an example radiograph with the annotated bounding boxes shown:

Prerequisites

Before building and executing the Docker image, make sure you have the following prerequisites installed on your machine:

Docker: Install Docker
Git: Install Git

Build and Execute the Docker Image

To build and execute the Docker image for the ROI Cavity Detection project, follow these steps:

Clone the repository:

git clone https://github.com/nitou2504/roi-cavity-detection.git

Navigate to the project directory:
```
cd roi-cavity-detection
```
Build the Docker image:
```
docker build -t roi-cavity-detection .
```
This may take some time as it installs the necessary dependencies.

Run the Docker Image

To run the Docker image for the ROI Cavity Detection project, follow these steps:

Create a folder named raw in the base directory of the project.
Inside the raw folder, place the periapical dental radiographs in JPG format. Each image should be accompanied by an XML file with the same name. The XML file should be formatted as Pascal VOC XML, providing information about the objects in the image, using the classes caries and no_caries. There is a sample radiograph JPG along with its accompanying XML file in the 'example' folder.

The folder structure and file format should look like this:
```
/path/to/base/dir
└── raw
    ├── image1.jpg
    ├── image1.xml
    ├── image2.jpg
    ├── image2.xml
    └── ...
```
Note: The XML files should follow the Pascal VOC XML format, which includes annotations for the objects in the image, such as bounding boxes and class labels.
Run the Docker image using the following command:
```
docker run -it -e BASE_DIR=. -v .:/app roi-cavity-detection
```
This command executes the Docker container in interactive mode inside the project directory roi-cavity-detection. Inside the container, the base directory will be available as the $BASE_DIR environment variable, and you will be in the container's shell.

Make sure to run the docker image inside the project directory, so it can access the python scripts and image files.

Note: The --entrypoint=/bin/bash flag can be used to start an interactive shell instead of running the default entrypoint specified in the Docker image, that runs the whole experiment starting from preprosessing the raw images. You can exit the container by typing exit in the shell.

License

This project is licensed under the GNU General Public License (GPL) version 3.0. See the LICENSE file for more details.