International Conference on Computer and Knowledge Engineering

Home / 13th International Conference on Computer and Knowledge Engineering

A Vision-Based Method for Human Activity Recognition Using Local Binary Pattern

Authors :

Babak Goodarzi¹ Reza Javidan² Mohammad Sadegh Rezaei³

1- Department of Computer Engineering and Information Technology, Shiraz University of Technology, Shiraz, Iran 2- Department of Computer Engineering and Information Technology, Shiraz University of Technology, Shiraz, Iran 3- Department of Computer Engineering and Information Technology, Shiraz University of Technology, Shiraz, Iran

Keywords :

Human Activity Recognition،Local Binary Pattern،Histogram of Oriented Optical Flow

Abstract :

Vision-Based Human activity recognition (HAR) plays an important role in various real-life applications. Recently, different solutions have been proposed and their performance is constantly improved. In this paper, we used a histogram of oriented optical flow (HOOF) that improved by extended and robust local binary pattern (LBP). Moreover, the effect of the local binary pattern with the HOOF method is also examined on different videos. The performance of the proposed method is evaluated by the KTH dataset, which contains six categories of human activities. The results showed that the accuracy of detection doesn’t change in waving, jogging and running activities; but for other activities including hand clapping, walking and boxing, the accuracy of detection become greater than 97 percent, which indicates the positive effect of using the features of local patterns.