Active 3D Scene Segmentation using Kinect.
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Active 3D Scene Segmentation Using Kinect
Author: Carlos Hernández Matas
This thesis takes as a starting point a segmentation framework for visual scenes based on Markov Random Fields. The approach integrates several cues and considers the scene in 3D, improving the hypotheses in an iterative manner.
In this thesis, we substitute the hardware previously used for a Kinect, a device capable of creating high quality depth maps of the environment.
The main interest of this work is to study the capabilities that the exploita- tion of said map can offer to the segmentation framework.
Specifically, we focus on the provision of an alternative attention mechanism to select fixations points for the segmentation, as well as two sets of cues. All of it based exclusively in the depth data retrieved from the Kinect.
The first set of cues is based in the detection of edges in said depth data, while the second set relies on the lack of depth information in specific areas.
Place, publisher, year, edition, pages
Trita-CSC-E, ISSN 1653-5715 ; 2011:103
IdentifiersURN: urn:nbn:se:kth:diva-130709OAI: oai:DiVA.org:kth-130709DiVA: diva2:654156
Master of Science - Systems, Control and Robotics
Bergström, NiklasEk, Carl Henrik