Maroš Bláha

© 2016 Maroš Bláha 

Analysis of semantic 3D city models (2016)

This thesis aims, on the one hand, to compare two different semantic reconstruction techniques in terms of geometric accuracy, and, on the other hand, to automatically extract 3D features (e.g. building corners, roof ridges, etc.) from a semantic city model. 

Student: Luca Miotti

Teaching


  • 103-0254-01L, Photogrammetry (5 ECTS credits), Spring Semester 2017, ETH Zurich
    • Responsible for the entire course
  • 103-0254-01L, Photogrammetry (5 ECTS credits), Spring Semester 2016, ETH Zurich
    • Lab supervision (implementation of basic multi-view algorithms)
  • 103-0274-00L, Image processing (3 ECTS credits), Spring Semester 2016, ETH Zürich
    • Lectures in the areas of image filtering and image restoration in the Fourier domain
  • 103-0254-01L, Photogrammetry (5 ECTS credits), Spring Semester 2015, ETH Zurich
    • Lab supervision (implementation of basic multi-view algorithms)
  • 103-0254-01L, Photogrammetry (5 ECTS credits), Spring Semester 2014, ETH Zurich
    • Lab supervision (implementation of basic multi-view algorithms)
  • 103-0798-00L, Geodetic Project Course (5 ECTS credits), Spring Semester 2013, ETH Zurich
    • Supervision of an UAV field campaign to monitor a landslide area with aerial images
  • 103-0254-01L, Photogrammetry (5 ECTS credits), Spring Semester 2013, ETH Zurich
    • Lab supervision (implementation of basic multi-view algorithms)

Supervised Student Theses

Semantic 3D modeling with UAV images (2015)

The topic of this thesis is the generation of a semantic 3D model (of a single house) based on UAV images, and the systematic, class-specific verification of the model against a ground truth model acquired with a laser scanner. 

Student: Nino Kracher

Reconstruction of a gravel pit with UAV images (2014)

To be able to accomplish surface and volume analysis, this thesis deals with the reconstruction of the gravel pit in Bülach (Switzerland) based on UAV imagery. The final model achieved a height accuracy of 0.1 m.

Student: Martin Baumann

Organization/Memberships



Geodetic project course (2013)

The purpose of this study was the detection of possible mass movements in a landslide area by combining two essentialy different techniques, namely gravimetry and photogrammetry.

Students: Christof Härri, Katharina Hiltebrand, Stefan Leuenberger, Jonatan Neukom, Tobias Nüssli, Aikaterini Tsampazi