Point Cloud Analytics and Application: Architectural Biometrics

Architectural Biometrics is an interdisciplinary project at the University of Rochester. In this project, we use lidar scanning along with computer vision and image processing technologies to address humanistic concerns in architectural history. More concretely, the Architectural Biometrics project aims at developing a computational tool that will allow researchers to compare incredibly detailed 3D scans of buildings built to a common design template. The humanistic goal of such comparison is to recognize the contributions of the engineers, tradespersons, and laborers who contributed to the on-site construction of each building, individuals who are often disenfranchised by attribution of the building to the architects alone. Project Website

  • Datasets
    • Two field sites: Canada and Turkey
    • Point clouds captured using lidar: FARO Focus 3D
    • Design drawings accessed from local library/archive
    • 19 railway stations in total (8 pairs and 1 trio)
field3 field3_pt metdesign1
  • Question of Interest
    • Are there fundamental structural differences between the buildings?
    • Do the constructed buildings differ from the design drawing?
    • Are there similarities among the deviation in individual buildings that share a common template specification?



  • Global Comparison Between Point Clouds
    met metheatmap ayra  ayraheatmap
    tur turheatmap demi demiheatmap
  • Global Comparison Between Point Cloud and Design Drawing
metplan elevation1  elevation2
  • Local Comparison Between Point Clouds
durakstairhm durakwall1 virtmp