A SOLUTION FOR HUMAN IDENTIFICATION BASED ON HOG AND SVM

Abstract

In surveillance systems, human detection and identification technology is being studied and applied to many areas with high reliability. This system can be applied in human resources management, security monitoring and so on. An effective solution for identifying and recognition human is important task in management, automated monitoring systems. In this article, the authors proposes an approach for identifying human identity based on the HOG descriptor for feature extraction and SVM for classification. The approach was evaluated on some datasets. In average, the accuracy rate of human identification achieved about 98.42%. Experimental results illustrated the proposed method can be applied to real applications.