Damselfly

Damselfly is an open-source, server-based photograph management system designed to index and search large, folder-based image collections. It focuses on fast search and keyword-tagging workflows, allowing users to manage their libraries through a web-based front-end. The software can be deployed on Windows, Linux, and macOS, and it is available as a Docker container for simplified installation and management.

Users interact with the system via a browser or a dedicated Electron.Net desktop client available for macOS, Windows, and Linux. The desktop application provides closer integration with the local operating system, enabling users to synchronize images from a selection basket on the server to a local folder. This workflow allows for the retrieval of images for editing in external software and subsequent re-syncing back to the server library.

Key features

• Local offline face detection, facial recognition, and object recognition
• Support for JPG, PNG, HEIC, TIFF, Webp, BMP, and RAW formats
• Full-text search with multi-phrase partial-word capabilities
• Non-destructive EXIF data updates using ExifTool
• Multi-user support with role-based entitlements and permissions
• Geolocation map display based on available image metadata
• Image reorganization tools to move, copy, or delete files
• Selection basket for exporting images or uploading to Wordpress
• Automated background indexing and thumbnail generation

The application is built with .NET 7, Blazor WebAssembly, and EFCore 7. It utilizes FaceONNX for facial recognition and YoloDotNet for object classification. The system is designed for high performance, capable of returning search results from a 500,000-image catalogue in less than one second. It includes a mechanism to exclude specific folders from scanning by adding a .nomedia file to the directory. The interface includes themes and a selection basket that can be kept private or shared with other registered users.

Damselfly is intended for photographers and collectors who require a centralized server to manage extensive image libraries across multiple devices without copying catalogues to local storage.