The Creepy Tracker eases prototyping with multiple commodity depth cameras. It automatically selects the best sensor to follow each person, handling occlusions and maximizing interaction space, while providing full-body tracking in scalable and extensible manners. It also keeps position and orientation of stationary interactive surfaces while offering continuously updated point-cloud user representations combining both depth and color data.
Here is an example scene from our Creepy Tracker project (Top: three users in front a wall display; Bottom: how the Creepy Tracker sees it):
The toolkit has four main components:
-
Sensor Client - A sensor client is connected to an individual depth sensor, and captures color, depth data and the body tracking model of every observed person in the tracked area.
-
Tracker Hub - The Tracker Hub component handles the unified model of the tracked area by combining the data streams from all sensor units.
-
Tracker Client for Unity3D - Set of Unity3D templates, with scripts and tools, that simplfies the development of new projects using the Creepy Tracker.
-
Surface Calibration - A standalone C# surface calibration tool.
Each component communicates with the others according to the following architechture:
Please cite this work using the following reference:
@inproceedings{Sousa:2017:CTT:3132272.3134113,
author = {Sousa, Maur\'{\i}cio and Mendes, Daniel and Anjos, Rafael Kuffner Dos and Medeiros, Daniel and Ferreira, Alfredo and Raposo, Alberto and Pereira, Jo\~{a}o Madeiras and Jorge, Joaquim},
title = {Creepy Tracker Toolkit for Context-aware Interfaces},
booktitle = {Proceedings of the 2017 ACM International Conference on Interactive Surfaces and Spaces},
series = {ISS '17},
year = {2017},
isbn = {978-1-4503-4691-7},
location = {Brighton, United Kingdom},
pages = {191--200},
numpages = {10},
url = {http://doi.acm.org/10.1145/3132272.3134113},
doi = {10.1145/3132272.3134113},
acmid = {3134113},
publisher = {ACM},
address = {New York, NY, USA},
keywords = {Context-aware Computing, Rapid-Prototyping, Toolkit},
}