Open-source photogrammetric tools for 3D urban modelling - A case study using mobile phone images

dc.contributorAalto-yliopistofi
dc.contributorAalto Universityen
dc.contributor.advisorIngman, Matias
dc.contributor.advisorVirtanen, Juho-Pekka
dc.contributor.authorKarki, Bijay
dc.contributor.schoolInsinööritieteiden korkeakoulufi
dc.contributor.supervisorVaaja, Matti
dc.date.accessioned2022-03-27T17:03:28Z
dc.date.available2022-03-27T17:03:28Z
dc.date.issued2022-03-21
dc.description.abstractInternet is flooded with images taken from mobile phone cameras. An estimate of more than 1.4 trillion photos was taken in the year 2021 and about 90% of which accounts for mobile phones. Technology has advanced a lot in the last decade when it comes to computer vision, image processing, and 3D reconstruction along with improvement in the camera hardware i.e. lens and sensor system in every new generation of phones. Several open-source software for 3D reconstruction are readily available which could offer cost-effective solutions as an alternative to commercial products. Moreover, the use of mobile phone images could even lower the cost of instruments. In this thesis, a 3D model of a façade on the southern end of the old Helsinki market hall in Kauppatori is created from the images taken from a mobile phone (Huawei P10) using one of the open-source photogrammetric tools i.e. Meshroom. Two sets of images were taken, each in auto focus and manual focus mode. A commercial tool i.e. Agisoft Metashape was used as a benchmark to create the 3D models using the same sets of images. All models were created in a complete default setting. The resulting mesh models were then compared to an existing TLS point cloud of the same structure. Cloud to mesh distance was computed for analyzing the accuracies of the models. Absolute RMSE values of both models generated from Meshroom were identical and less than 8 cm. However, visual observation clearly shows the models had differences in absolute and local accuracies. Several cross sections were analyzed to access the local accuracies within the models. The models tend to warp away from the point cloud towards the outer edges of the façade. Since the study was also intended to encourage beginners with a DIY approach, data acquisition and processing were made as simple as possible. So, there are several rooms for improvement. The results demonstrate a tremendous possibility of open-source photogrammetric tools and mobile phone images as a cost-effective solution for 3D modelling applications.en
dc.format.extent42+8
dc.format.mimetypeapplication/pdfen
dc.identifier.urihttps://aaltodoc.aalto.fi/handle/123456789/113659
dc.identifier.urnURN:NBN:fi:aalto-202203272541
dc.language.isoenen
dc.programmeMaster's Programme in Geoinformatics (GIS)fi
dc.programme.majorfi
dc.programme.mcodefi
dc.subject.keywordmobile phone photogrammetryen
dc.subject.keywordfacade modellingen
dc.subject.keywordmeshroomen
dc.subject.keywordopen-source photogrammetryen
dc.titleOpen-source photogrammetric tools for 3D urban modelling - A case study using mobile phone imagesen
dc.typeG2 Pro gradu, diplomityöfi
dc.type.ontasotMaster's thesisen
dc.type.ontasotDiplomityöfi
local.aalto.electroniconlyyes
local.aalto.openaccessyes

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
master_Karki_Bijay_2022.pdf
Size:
3.34 MB
Format:
Adobe Portable Document Format