\section{Related Work}
\label{sec:related}

%The related work for this paper is divided into two main sections: using
%biometrics for identification and advances in skeleton mapping.
%
%\paragraph{Biometrics}

%One way to view different biometrics are how intrusive they are.  More
%intrusive biometrics include fingerprints, irises, and DNA; less intrusive are
%faces, speech, and gait~\cite{phillips98}.  Generally, there is a tradeoff
%between the level of user cooperation required and the accuracy that can be
%attained~\cite{liu01,bio-survey}.

A biometric is a distinguishable trait that can be measured and used for
identification.  There are many different possible biometrics.  
Biometrics can be divided into physiological and behavioral traits.
Physiological traits include faces, fingerprints, and irises; speech and gait
are behavioral.  Faces and gait are the most relevant biometrics for this paper
as they both can be collected passively and involve image processing.

Approaches to gait recognition typicaly fall into two categories: silhouette-based and
model-based.  Silhouette-based techniques recognize gaits from a binary
representation of the silhouette as extracted from each image, while
model-based techniques fit a 3-D model to the silhouette to better track
motion~\cite{gait-survey}. Some model-based techniques use features of the
generated model to aid in gait recognition, but they are severely limited by
the difficulty in generating a 3-D model from a 2-D
image~\cite{gait-body1,gait-body2}.  Furthermore, behavioral traits typically
are more characteristic as opposed to unique, and are subject to change with
time and based on the observed activity~\cite{gait-survey2}.

Face recognition, as a physiological biometric, is a more static feature.  Face
recognition can be broken down into three parts: face detection, feature
extraction, and classification; these three parts are studied both individually
and combined~\cite{face-survey}.  The Xbox 360 uses face recognition with the
Kinect as part of its user recognition algorithm, in addition to the height
inferred from the Kinect~\cite{kinect-identity}.

In this paper, we propose using skeleton measurements as a biometric separate
from face and gait biometrics.  According to Jain~\etal{}~\cite{bio-survey}, a
proper biometric has the following characteristics: \first \emph{universality},
everyone has it, \second \emph{uniqueness}, it should be different between any
two people, \third \emph{permanence}, it does not vary with time, and \fourth
\emph{collectability}, it is measurable. Universality, permanence, and
collectability are easily met with skeletons (skeletal changes that occur with
age happen gradually).  We discuss how uniqueness is met in detail in
\xref{sec:uniqueness}.

By using skeleton as a biometric for recognition, we can formulate skeleton
recognition in a similar way as we can face recognition.  The equivalent parts
would be figure detection, skeleton fitting, and classification. Figure
detection and skeleton fitting respectively map to silhouette extraction and
model fitting in gait detection, but as previously noted, they are severely
limited.  However, Zhao~\etal~\cite{zhao20063d} perform gait recognition in 3-D
using multiple cameras.  By moving to 3-D, many of the problems related to
silhouette extraction and model fitting are removed.  Additionally we can take
advantage of the wealth of research relating to 3-D motion
capture~\cite{mocap-survey}. 
%Specifically, range cameras offer real-time depth
%imaging, and 
The Kinect~\cite{kinect} is a widely available 3-D sensor (also known as a
range camera) with a low price point, that has been leveraged to improve gait
recognition~\cite{munsell:eccv12,hoffman:btas12}.  Figure detection and
skeleton fitting have also been studied in motion capture, mapping to region of
interest detectors and human body part identification or pose estimation
respectively in this
context~\cite{plagemann:icra10,ganapathi:cvpr10,shotton:cvpr11}.  Furthermore,
OpenNI~\cite{openni} and the Kinect for Windows SDK~\cite{kinect-sdk} are two
systems that perform figure detection and skeleton fitting for the Kinect.
Given the maturity of the solutions, we will use existing implementations of
figure detection and skeleton fitting.  Therefore this paper will focus
primarily on the classification part of skeleton recognition.


%a person from an image to measure gait, but can also be measured from floor
%sensors or wearable sensors~\cite{gait-survey}.  
%
%Face recognition can be broken
%down into three parts: face detection, feature extraction, and identification;
%these three parts are studied both individually and together~\cite{face-survey}.
%We will compare skeleton recognition to face recognition in this paper for the
%following two reasons.  
%First, behavioral traits typically are more
%characteristic as opposed to unique for identification beyond a certain
%scale~\cite{seven-issues}, 
%and second, face recognition is more widely studied
%and accepted as a means of accurate identification~\cite{face-survey}.
%
%\paragraph{Skeleton mapping}
%
%but algorithms similar to gait
%recognition have been developed for range cameras as well~\cite{gomez:hgbu11}.
%
%also including at least one example incorporating
%a range camera~\cite{gordon91}.