Investigating effective mechanisms for crowdsourcing photogrammetry-based 3D cultural heritage via collective behaviour analysis

  • Danzhao CHENG

Student thesis: PhD Thesis


Establishing openly accessible interconnected databases of 3D cultural heritage objects holds significant potential to foster awareness, education, and appreciation of cultural heritage among a broader global audience. These platforms also serve as rich sources of inspiration for the creative use of cultural heritage. Nevertheless, traditional digitisation means, typically involving experts scanning, are resource-demanding and not scalable. Mass photogrammetry can utilise the ubiquity of photographic devices and the prevalence of photographic habits, together with the voluntary attitudes of crowds, to achieve the goal of obtaining large-scale images of target objects, even beyond the reach of formal cultural institutions. However, the challenge remains in understanding how to effectively outsource effort-intensive image acquisition tasks to non-expert volunteers to generate quality 3D reconstructions.
The present research investigates effective mechanisms for crowdsourcing photogrammetry-based 3D cultural heritage. It starts by identifying a generic framework and examines the potential of synchronous offline collaboration in mapping the collective tasks to enhance the quality of the 3D model. To mitigate coordination costs, Various social computing techniques are employed to seamlessly integrate collaborative structures into situated crowdsourcing. To validate its viability, this situated open collaboration platform undergoes a comprehensive field study involving different crowdsourcing scenarios. The cross-evaluations affirm its effectiveness, as it adeptly facilitates asynchronous collaborations among situated volunteers, stimulating heightened levels of self-selective participation and high-quality spontaneous contributions. This approach’s systematic optimisation of task allocation and distribution substantially enhances the efficiency and collective outcomes of mass photogrammetry. This research extends its scope to explore the spatio-temporal crowdsourcing dynamics within collaborative mass photogrammetry. The follow-up study employs reflective assessments with quantitative inquiries to explore volunteers' prosocial motivation, assessing underlying correlates and structures. The results demonstrate adequate reliability and statistically validate empirical findings. It pioneers the understanding of prosocial motivation in collaborative crowdsourcing, providing practical implications for incentive mechanisms to leverage collective performance in future task-oriented collaborative voluntary activities. Finally, this study synthesises the empirical and theoretical findings, presenting a systematic approach for crowdsourcing 3D cultural heritage objects via mass photogrammetry. This study serves as a complementary extension of current crowdsourcing research, aimed at enhancing data quality and collective productivity. Moreover, the investigation derives concrete design implications for situated collaborative crowdsourcing, enlightening its potential in connecting serendipitously available volunteers for contextually-relevant information. Furthermore, this research contributes pragmatic insights into volunteer-based technology use and the significance of crowd-generated content contribution in the digital heritage domain.
Date of AwardOct 2023
Original languageEnglish
Awarding Institution
  • University of Nottingham
SupervisorEugene Chng (Supervisor) & Xinwei Wang (Supervisor)


  • Crowdsourcing
  • Collective Behaviour
  • Mass Photogrammetry
  • Cultural Heritage
  • CSCW
  • Human-centered Computing
  • Collaborative and Social Computing
  • Applied Computing
  • Digital Archives
  • Empirical Analysis

Cite this