The tool combines web-based and AR technologies to enhance the presentation and engagement with fashion heritage objects. 3D models of heritage elements (geometry and texture) from laser scanning or photogrammetric acquisitions are optimised for web-based and AR visualisation. This app enables users to explore and interact with fashion artefacts from various eras through web browsers and AR devices. Users can virtually visualise the intricate details of garments, virtually "try on" historical attire using AR, and gain insights into the context and history of each piece. The application is, in particular, designed for the virtual fitting of different-shaped historical hats

The virtual try-on app has been developed as a web application to maximise its deployment and reuse possibilities, ensuring independence from specific platforms and operating systems. FBK utilised Snap Lens Studio and the Camera Kit Web SDK from Snap Inc. to create the app, which is hosted on EFHA’s servers. The deployed app is fully compliant with GDPR, and it doesn’t store any photos or user information.

The application includes seven 3D models of historical hats and one pair of glasses. All the 3D models have been published to Europeana via the EFHA aggregator.  In the first phases of the project, several 3D reality-based acquisition techniques have been tested to verify the achievable quality of the models. The selected garments were surveyed with both active and passive sensors (portal LiDAR solutions and close-range photogrammetry), and the results were evaluated in terms of completeness, density, and quality of the derived textures. Although the results were comparable in terms of geometry, the superior quality of the textures generated with photogrammetric techniques led to the selection of these results for further processing. 





 | 

The 3D photogrammetric models of the fashion garments have been meticulously optimised for reuse in the AR app. This process involved both geometry and texture optimisation (faces decimation and texture enhancement) to reduce file sizes and improve performance while maintaining a high level of visual fidelity. 

These optimisations ensure smoother rendering and faster loading times, enhancing the overall user experience. Additionally, the models were adapted to be compatible with the AR framework, allowing for seamless integration and interaction within the app.