Interactive Geometry-Aware Segmentation for the Decomposition of Kaleidoscopic Images
Mirror systems have recently emerged as an alternative low-cost multi-view imaging solution. The use of these
systems critically depends on the ability to compute the background of a multiply mirrored object. The images
taken in such systems show a fractured, patterned view, making edge-guided segmentation difficult. Further,
global illumination and light attenuation due to the mirrors make standard segmentation techniques fail.
We therefore propose a system that allows a user to do the segmentation manually. We provide convenient tools
that enable an interactive segmentation of kaleidoscopic images containing three-dimensional objects. Hereby, we
explore suitable interaction and visualization schemes to guide the user. To achieve interactivity, we employ the
GPU in all stages of the application, such as 2D/3D rendering as well as segmentation.
Projects

A Kaleidoscopic Approach to Surround Geometry and Reflectance Acquisition
In: CVPR Workshop on Computational Cameras and Displays (CCD), 2012
Abstract
We describe a system for acquiring reflectance fields
of objects without moving parts and without a massively
parallel hardware setup. Our system consists of a set of
planar mirrors which serve to multiply a single camera and
a single projector into a multitude of virtual counterparts.
Using this arrangement, we can acquire reflectance fields
with an average angular sampling rate of about 120+
view/light pairs per surface point. The mirror system allows
for freely programmable illumination with full directional
coverage. We employ this setup to realize a 3D acquisition
system that employs structured illumination to capture the
unknown object geometry, in addition to dense reflectance
sampling. On the software side, we combine state-of-the-art
3D reconstruction algorithms with a reflectance sharing
technique based on non-negative matrix factorization
in order to reconstruct a joint model of geometry and
reflectance. We demonstrate for a number of test scenes
that the kaleidoscopic approach can acquire complex
reflectance properties faithfully. The main limitation is
that the multiplexing approach limits the attainable spatial
resolution, trading it off for improved directional coverage.
Video Bibtex
@InProceedings{Ihrke12,
author = {Ivo Ihrke and Ilya Reshetouski and Alkhazur Manakov and Art Tevs and Michael Wand and Hans-Peter Seidel},
title = "{A Kaleidoscopic Approach to Surround Geometry and Reflectance Acquisition}",
booktitle = {Proceedings of IEEE International Workshop on Computational Cameras and Displays },
pages = "1--8",
year = {2012},
}
author = {Ivo Ihrke and Ilya Reshetouski and Alkhazur Manakov and Art Tevs and Michael Wand and Hans-Peter Seidel},
title = "{A Kaleidoscopic Approach to Surround Geometry and Reflectance Acquisition}",
booktitle = {Proceedings of IEEE International Workshop on Computational Cameras and Displays },
pages = "1--8",
year = {2012},
}