Abstract:
Privacy of images is a growing concern in this day and age. In recent times 3D
secure maps offered a solution to hide details but 3D maps the geometry of the
scene which leads again to privacy violation. Objects can be detected in the scene
and that leads to localization of the objects present in the scene. In this thesis
we aim to evaluate these privacy features in a 3D map. Our analysis is to detect
objects if found in any map, evaluate its localization and how many instances of
same objects can be there in this revealing process. Our Analysis will eventually
help in designing more robust privacy features in order to protect privacy.
