Human Gait Recognition Usingn Preprocessing and Classification Techniques

samer kais Jameel

Abstract


Biometric recognition systems are of concern to many researchers, as they try to overcome the obstacles and factors that weaken these systems, such as problems of obtaining images i.e. the resolution or the object does not appear completely...Etc. In this paper, reliance is placed on the movement of the object, where the human can be distinguished through its gait. Lest of losing some of the features that may impair the ability of system to recognize people, we have proposed to use of all data recorded by the Kinect sensor without extracting the features from those data as in previous researches, where, coordinates of 20 points were recorded for each person, that produced 8404 constraints, they were walking at different speeds and directions, and for different genders and ages. Furthermore, pre-Processing techniques are used and measure its effects on the efficiency of the system through tested it on six types of classifiers. The proposed approach achieved a noteworthy recognition rate, which is reached 91% without using descriptors.



DOI: http://doi.org/10.11591/ijece.v10i3.pp%25p
Total views : 30 times

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.