A Statistical Learning Approach to Image Processing

抄録

In this paper proposes an object detection method that uses Histograms of Oriented Gradients (HOG) features using two-stage boosting. There has been done many research works in recent years on statistical training methods and object detection methods that combine low-level features obtained. However, in our proposed approach, low-level HOG features are combined by using Real AdaBoost to automatically generate features. In this way, it is possible to capture a shape of symmetry and edge continuity, which single HOG features cannot do, so highly accurate detection is realized. In 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 the HOG features appearance for detecting a pedestrian.