Human movement Recognition using Euclidean Distance: A tricky approach
Joveria Javed, Hashim Yasin, Syed Faisal Ali · 2010 3rd International Congress on Image and Signal Processing · 2010
In this research paper, the Euclidean Distance based Recognition (EDR) methodology has been proposed and then different classifiers are applied to recognize and classify different human movements. The classifiers used in this research paper are: Mahalanobis Distance (MD) classifier; Quadratic Discriminate Analysis (QDA) classifier, Linear Discriminate Analysis (LDA) classifier and Fuzzy K-Nearest Neighbor (FKNN) classifier. Five different human movements namely sit down, jump up, arm down, bend down and kick front performed by different subjects at different timings have been recorded. These different human movements have been detected and recognized by means of Euclidean Distance based Recognition (EDR) methodology and Hu Moment Invariant based Recognition (HMIR) methodology. At the end, the performance of Euclidean Distance based Recognition (EDR) methodology is compared with that of Hu Moment Invariant based Recognition (HMIR) methodology in terms of accuracy ratio, memory capacity and total average time.