Farzad, Mohammad Reza (2016-05). Interior Close-Range Digital Photogrammetry for an Operational Building. Master's Thesis. Thesis uri icon

abstract

  • Laser scanning technology has been well received by the industry practitioners who want to create a 3-dimensional (3D) computer model of existing buildings and structures. However, most 3D laser scanners are expensive and need a special training to be utilized. It also takes a significant amount of time to convert point clouds collected from the laser scanner into a 3D computer model. Photogrammetry technology may provide an alternative solution to those who want to pick up the 3D model of an existing building. Since it uses snapshot photos of the target object taken at multiple locations, one can create a semi-realistic 3D model of a building cost-effectively without spending their budget for expensive laser scanners. Basically, it is designed to pick up a 3D model of an object using photos taken from multiple locations outside an object. So, one may speculate if this algorithm would work if someone wants to pick up the 3D model inside the building. If it is possible, then others may be wondering how long it would take to pick up the 3D model of a building's interior, or if the 3D model would be accurate enough for facilities management or other activities sought by construction managers. This study evaluates the practicality of photogrammetry technology in creating a 3D model of a building's interior. A commercial photogrammetry-based 3D modeling application was used to test if one can create the 3D model of the building interior, how long it takes to pick up a 3D model, and how accurate the 3D model would be. The Francis Hall building located on the Texas A&M University campus in College Station was chosen for this test. As many as 4,940 photos were taken at multiple locations inside the building. Autodesk Recap 360 was then used to create the 3D model of a unit space inside the building. Unity 3D was used to combine multiple unit 3D models into a 3D model of the entire floor of the Francis Hall. This test demonstrates that it is possible to create a unit 3D model from building's interior space using photographs. It took 47.95 minutes in average to take photos in one room, and 647.05 minutes to produce its 3D model. In total, it took 14,469 minutes to produce a 3D model of the entire first floor of the Francis Hall. Average tolerance between the real measurements and the dimension of the 3D model from photos is about 0.83 percent.
  • Laser scanning technology has been well received by the industry practitioners who want to create a 3-dimensional (3D) computer model of existing buildings and structures. However, most 3D laser scanners are expensive and need a special training to be utilized. It also takes a significant amount of time to convert point clouds collected from the laser scanner into a 3D computer model.

    Photogrammetry technology may provide an alternative solution to those who want to pick up the 3D model of an existing building. Since it uses snapshot photos of the target object taken at multiple locations, one can create a semi-realistic 3D model of a building cost-effectively without spending their budget for expensive laser scanners.

    Basically, it is designed to pick up a 3D model of an object using photos taken from multiple locations outside an object. So, one may speculate if this algorithm would work if someone wants to pick up the 3D model inside the building. If it is possible, then others may be wondering how long it would take to pick up the 3D model of a building's interior, or if the 3D model would be accurate enough for facilities management or other activities sought by construction managers.

    This study evaluates the practicality of photogrammetry technology in creating a 3D model of a building's interior. A commercial photogrammetry-based 3D modeling application was used to test if one can create the 3D model of the building interior, how long it takes to pick up a 3D model, and how accurate the 3D model would be. The Francis Hall building located on the Texas A&M University campus in College Station was chosen for this test. As many as 4,940 photos were taken at multiple locations inside the building. Autodesk Recap 360 was then used to create the 3D model of a unit space inside the building. Unity 3D was used to combine multiple unit 3D models into a 3D model of the entire floor of the Francis Hall.

    This test demonstrates that it is possible to create a unit 3D model from building's interior space using photographs. It took 47.95 minutes in average to take photos in one room, and 647.05 minutes to produce its 3D model. In total, it took 14,469 minutes to produce a 3D model of the entire first floor of the Francis Hall. Average tolerance between the real measurements and the dimension of the 3D model from photos is about 0.83 percent.

publication date

  • May 2016