Document scanner app with edge detection and cropping functionality.
The app scans a document image, detects the edges and applies a magnifier to crop the document.
- On-device machine learning for document scanning
- Edge detection capability
- Use of OpenCV edge detection library and ffi to connect native C++ to Flutter with the help of Dart Isolates.
- Ability to crop and edit images
- Ability to download the cropped image
Although the project is a simple app in terms of its frontend capacity, the project was still implemented as though it was going to grow into an enterprise app using the MVVM approach.
- State Management was done with setState (with a growth route towards use of Riverpod for SM and DI)
- File/Folder structure ensures large codebase ease of maintenance
The project depends on a plugin I had to develop for the OpenCV integration to work perfectly. This plugin has large files that could not be uploaded to Github, hence, I had to use the local version of the Plugin. Explanation on how to get the files/folders for the Plugin is explained in the Plugin's Readme. After this, the plugin path can be changed in the pubspec.yaml file to the local directory of the Plugin on your machine.
- The web and desktop environments were not prioritized for this submission.
- CI / CD was not prioritized for this submission.
- Flavors were not prioritized for this submission.
- Analytics and Error Reporting was not prioritized for this submission.
- run
git clone https://github.com/IdrisAdeyemi01/docu_edges.gitin your terminal - navigate to the directory containing the cloned project (docu_edges)
- open in any code editor of your choice e.g Android studio, VSCode etc
- Complete setup as mentioned in the Build Limitation section
- connect your Android device
- run
flutter runto test in your connected device
- You can also simply install the app's apk to text. Docu Edges
Built with ❤️ in Flutter. Powered by OpenCV through Dart FFI.