A multi-lod approach for privacy enabled 3D building model

The primary objective of this study is to investigate present privacy issues in 3D city models. Along with the developments on 3D city visualization technologies such as Google Street View, Open Street View, municipal city information system software; locational information has been merged with the high resolution visual imageries. Accordingly, personal information became visible and accessible by public easily and it created a risk factor for societies on security of personal information. In light of this, the problems have been examined and solutions have been investigated in the thesis study. Current studies regarding privacy problems on visual city models are not sufficient to correspond the personal concerns of visible information. Based on image blurring and obfuscation techniques, privacy preserving practices today cause the lack of data as a result of focusing on single objects. It has been required to search and improve another approaches in which it is possible to hide personal information while keeping the other details being necessary for the models such as architectural appearances, commercial blocks. In the implementation part of the study, a sample 3D building model has been designed in different level of details and used in order to illustrate the main concepts of modelling through privacy matters. For this purpose, existing buildings in the model have been divided into their floors and remodeled in every abstraction level. Subsequently, three decomposition building models have been provided which are consisting of three different detail levels for each floor and the privacy levels as well. The paper proposes to solve the privacy problems on 3D city models by producing 3D models in many detail levels at the same time and generating decomposition models. Therefore, to preserve personal information on buildings can be possible without any lack of the other visual information through the instrument of recombined building models in the direction of needs.