Researchers Develop Modular 4D World Web-Visualisation Software Framework

Strategic Insights -

Researchers at the Friedrich-Schiller-Universitat Jena have developed a modular 4D world web-visualisation software framework for both mobile and desktop devices. This framework aims to address the challenge of obtaining and compiling 4D world models at a large scale. The research proposes a data retrieval pipeline and geolocalisation, as well as 3D data processing to blend 3D assets with digital elevation models for seamless 3D environments.

Key Takeaways:

  • The researchers have developed a modular 4D world web-visualisation software framework for both mobile and desktop devices, which has been in development since 2016.
  • The framework aims to address the challenge of obtaining and compiling 4D world models at a large scale.
  • The research proposes a data retrieval pipeline and geolocalisation, as well as 3D data processing to blend 3D assets with digital elevation models for seamless 3D environments.
  • The study has validated different approaches for rough and fine level geolocalisation using Large Language and Vision Language Models, as well as key point matchers.
  • The research has also included the preprocessing of 3D assets to generate 3D environments.
  • The study has used ground-truth material taken from datasets of 75,000 3D mesh models and 1.5 million images.
  • The researchers have shown results from the validation of the pipelines, including the use of Large Language and Vision Language Models and key point matchers.
  • The study has been published in The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences (2025,XLVIII-M-9-2025):1043-1050.

Statistics:

  • 75,000 3D mesh models have been used in the study.
  • 1.5 million images have been used in the study.
  • The research has taken 9 years in development, starting from 2016.
  • The study has validated different approaches for rough and fine level geolocalisation.
  • The researchers have used Large Language and Vision Language Models, as well as key point matchers in the study.

Sources:

  • 4D World Viewers as Multi-user Content Management Systems. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2025,XLVIII-M-9-2025():1043-1050.
  • The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - http://www.isprs.org/publications/archives.aspx.
  • Information Technology Newsweekly, October 21, 2025; p 517.