Simple native application to detect multi-object durian leaf disease.

LeafEye is a native Windows desktop application developed using C++/WinRT and WinUI 3 (Windows App SDK). It utilizes a deep learning model to perform multi-object detection on durian leaves, classifying various diseases locally. The application relies on an offline-first architecture powered by ObjectBox to manage user data, scan histories, and detection results without requiring an active internet connection.

• Deep Learning Integration: Processes image data to detect and classify multiple bounding-box instances of leaf diseases in a single execution.
• Native Frontend: Built on WinUI 3 with C++/WinRT projections. UI updates are handled reactively using IObservableVector combined with INotifyPropertyChanged for minimal rendering overhead.
• Local Storage: Integrates ObjectBox (C++ API) for NoSQL data persistence. Database transactions are isolated on background worker threads to maintain main-thread fluidity.
• Concurrency: Heavily utilizes C++ coroutines (co_await, winrt::fire_and_forget) and the DispatcherQueue to handle asynchronous data fetching and safe UI thread marshalling.
• Data Models: Relational schema handling User credentials, Profile metadata, batch Scan History, and individual File History (tracking confidence scores and disease IDs).
• Application Telemetry: Implements a custom, mutex-locked file logger to capture runtime execution states and global unhandled exceptions to %LOCALAPPDATA%.

• UI Framework: WinUI 3 / Windows App SDK
• Language: C++17 / C++/WinRT
• Database: ObjectBox
• Markup: XAML

To compile and run this application locally, the following development environment is required:

• Visual Studio 2022 (v17.0 or later)
• "Desktop development with C++" workload
• Windows App SDK C++ Templates component
• Windows 10 SDK (10.0.19041.0 or later)

1. Clone the repository:

   git clone https://github.com/daberpro/LeafEye.git