Human Classification Using Gait Features

Abstract

Abstract. Gait exhibits several advantages with respect to other biometrics features: acquisition can be performed through cheap technology, at a distance and without people collaboration. In this paper we perform gait analysis using skeletal data provided by the Microsoft Kinect sensor. We defined a rich set of physical and behavioral features aiming at identifying the more relevant parameters for gait description. Using SVM we showed that a limited set of behavioral features related to the movements of head, elbows and knees is a very effective tool for gait characterization and people recognition. In particular, our experimental results shows that it is possible to achieve 96% classification accuracy when discriminating a group of 20 people.Keywords: Gait characterization, Gait analysis, Kinect, Support Vector Machine Introduction and Related WorkBiometrics is the science that studies the human characteristics for anthropometry research, people identification, access control and many more. Biometric features are measurable data classified as physical or behavioral [1]. The former are related to the body and its shape. Some examples are face, hand, iris, retina and fingerprint. Behavioral characteristics are associated to particular human action, for instance handwriting and walking. Automatic recognition systems are often expensive, intrusive and require the cooperation of the subject during the acquisition. The latter cannot be always guaranteed, for instance in a video surveillance context. In this case, it is useful to recognize people through biometric parameters that can be captured at a distance and without the collaboration of the person, such as gait [2].Gait analysis finds interest in video surveillance systems [3,4] and forensics science [5,6]. Furthermore, many applications analyze the gait in order to discover pathologies of the body movement [7], rehabilitation therapy [8], identify the fall risk in elderly population in order to assess the frailty syndrome [9,10]. All these applications are based on the analysis of video and 2D images. Images and videos are processed in order to collect gait parameters applying both model-based approaches, using the definition of a 3D model of the body in movement [11][12][13], or by model-free approaches, that process the silhouette of a walking person [14]. In this paper we implemented a model-based approach using the Microsoft Kinect sensor. Kinect is more that a simple RGB camera, since it is equipped with a depth sensor providing 3D information related to the movements of body joints. The 3D data are more precise compared to the 2D information extracted from images but, on the contrary, the depth sensor, based on the infrared rays, does not work in outside environment and the depth range is quite limited. In literature exists some applications that exploit Kinect 3D data for people recognition and classification. Preis et al. [15] used only anthropometric features, such as height, length of limbs, stride length and speed, for gait characterization. They te...

Kinect-based gait analysis for automatic frailty syndrome assessment

“…In our implementation we repeat the same steps on both right and left strides. Stride detection is based on the ankles joints whose estimated positions turn to be stabler than the feet joints (see our previous analysis in [16,17]). …”

“…In [16,17] it is shown that gait features extracted from Kinect can be used as a biometric signature. Gabel et al [18] have performed full body gait analysis for continuous gait monitoring at home.…”

Human Gait Recognition Using Skeleton Joint Coordinates with Orthogonal Least Square and Locally Linear Embedded Techniques

“…Thus, in order to reduce the computational cost as well as to acquire accurate model fitting, many researches related to algorithms for modeling and tracking body part, such as primitive shapes and stick figures have been carried out [7], [8], [18]- [23]. However, recent researches attempt to overcome the drawbacks by applying a new generation of image acquisition device, known as Kinect Sensors for human gait recognition purpose [24]- [29]. Compared to a standard video camera, Kinect generates model fitting of stick-figure model accurately and further tracks human without involving complex processes for distance between 0.8 m to 4 m [30].…”

“…Researchers developed feature extraction technique that involved both static and dynamic features. As a result, high recognition performance were obtained based on findings reported based on static features [27], [30]- [33] as well as fusion of static and dynamic features [26], [28], [29], [34], [35]. The worst recognition performance reported using dynamic features were findings from [26], [27], [36], [37].…”

Folk Dance Pattern Recognition Over Depth Images Acquired via Kinect Sensor

“…In the work (Gianaria et al, 2014), gait analysis is performed using skeletal data provided by Microsoft Kinect sensors and a set of physical and behavioural features is defined, in order to identify the more relevant parameters for gait description. The aim of this work is the gait characterization and people recognition using SVM classification.…”