1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
\section{Introduction}
\label{sec:intro}
Person identification has become a valuable asset, whether for means of
authentication, personalization, or other applications. Previous work revolves
around either physiological biometrics, such as face recognition, or behavioral
biometrics such as voice or gait recognition. In this paper, we propose using
skeletal measurements as a new physiological biometric for recognition.
In recent years, advances in range cameras have given us access to increasingly
accurate real-time depth imaging. Furthermore, the low-cost and widely
available Kinect~\cite{kinect} has brought range imaging to the masses. In
parallel, the automatic detection of body parts from depth images has led to
real-time skeleton mapping. As the resolution and accuracy of range cameras
improve, so will the accuracy and precision of skeleton mapping algorithms.
In this paper we show that skeleton mapping is accurate and unique enough in
individuals to be used for person recognition. First, we show that ground
truth skeleton measurements can uniquely identify a person. Second, we show
how the accuracy of skeleton recognition decreases as simulated error
increases. Third, we collect skeleton data with a Kinect in an uncontrolled
setting and we apply preprocessing and classification algorithms to this
dataset. We evaluate the performance of skeleton recognition with varying
group size and compare it to face recognition.
Much of the prior work in person recognition focuses on data gathered from
other sensors, such as face recognition with color cameras and voice
recognition with microphones. In the realm of depth imaging, most of the work
surrounds behavioral recognition, continuing work in gait recognition. The
Xbox 360~\cite{} does use the height inferred from the Kinect as part of its
user identification algorithm, albeit in addition to other attributes including
face recognition.
|
