Human tracking method based on improved HOG+Real AdaBoost

This paper proposes an object detection method that uses Histograms of Oriented Gradients (HOG) features using boosting algorithm. There has been done many research works in late years on statistical learning methods and object detection methods that associate low level of features obtained. However the proposed approach, low level of HOG features are associated by using Real AdaBoost to continuously achieve features. In this wise, it is possible to capture a shape of edge continuity, which single HOG features can´t do, so highly accuracy detection is realized. This paper, to evaluate the effectiveness of the proposed method, three different experiments with different patterns are conducted for detecting humans. Moreover, a boosting classifier is used to represent the co-occurrence of HOG features appearance for detecting a human.