Modified kNN Algorithm for Improved Recognition Accuracy of Biometrics System Based on Gait

The k-nearest neighbors classifier is one of the most frequently used. It has several interesting properties, though it cannot be used in utilitarian biometric systems. This paper proposes the modification of kNN algorithm which ensures correct work even in the case of an attempt at access to the system by unauthorized people. Work of the algorithm was tested on data presenting ground reaction forces (GRF), generated during human’s walk, obtained from measurements carried out on over 140 people. Differences between particular strides were determined with the Dynamic Time Warping (DTW). The recognition precision obtained was as high as 98% of biometric samples.

[1]  Nicolás García-Pedrajas,et al.  Boosting k-nearest neighbor classifier by means of input space projection , 2009, Expert Syst. Appl..

[2]  S. P. Moustakidis,et al.  Feature extraction based on a fuzzy complementary criterion for gait recognition using GRF signals , 2009, 2009 17th Mediterranean Conference on Control and Automation.

[3]  Gregory D. Abowd,et al.  The smart floor: a mechanism for natural user identification and tracking , 2000, CHI Extended Abstracts.

[4]  J. Cutting,et al.  Recognizing friends by their walk: Gait perception without familiarity cues , 1977 .

[5]  Jacek M. Zurada,et al.  Artificial Intelligence and Soft Computing, 10th International Conference, ICAISC 2010, Zakopane, Poland, June 13-17, 2010, Part I , 2010, International Conference on Artificial Intelligence and Soft Computing.

[6]  Tingting Mu,et al.  Gait recognition: highly unique dynamic plantar pressure patterns among 104 individuals , 2012, Journal of The Royal Society Interface.

[7]  Jaeseok Yun,et al.  User Identification Using Gait Patterns on UbiFloorII , 2011, Sensors.

[8]  Marcin Derlatka,et al.  Human Gait Recognition Based on Signals from Two Force Plates , 2012, ICAISC.

[9]  Einar Snekkenes,et al.  User authentication based on foot motion , 2011, Signal Image Video Process..

[10]  Sung-Hyuk Cha,et al.  Performance of a long-text-input keystroke biometric authentication system using an improved k-nearest-neighbor classification method , 2010, 2010 Fourth IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS).

[11]  Yu-Tzu Lin,et al.  Human recognition based on kinematics and kinetics of gait , 2011 .

[12]  Maricor N. Soriano,et al.  Compact time-independent pattern representation of entire human gait cycle for tracking of gait irregularities , 2010, Pattern Recognit. Lett..