Gait Recognition and Classification Using Random Forest Algorithm

M. Hema, K. Babulu and N. Balaji

In present competitive world technology have a huge impact on every aspect of life. This technology is getting helpful in every field, if we talk about a person identification which is a great task in present days have evolved number of biometric techniques such as face, iris, finger print, voice etc. but beyond these all Gait based human identification technique is best in use where a person can be identified with his silhouette image. In this paper we see how Random Forest algorithm is efficient than the SVM classifier technique. We view a person in 11 different angles where every view is 18 degrees varying (0-180 degrees) and a person will be in three cases such as with bag, coat and normal walk. This is one of the cases in CASIA-B. This database is used for the experimentation in order to obtain the high accuracy of gait recognition and we compare the obtained classifiers results for efficiency.

Volume 11 | Issue 2

Pages: 281-289