An extended HOG model: SCHOG for human hand detection

It is crucial to get human hand information for hand gesture recognition tasks. However, at present, people can not still get a perfect hand segmentation or localize hand accurately especially under complex conditions. Therefore, it is necessary to develop robust and effective methods for detecting human hand accurately. In this paper, we propose a new method for hand detection. We present an extended histogram of oriented gradients (HOG) model: skin color histogram of oriented gradients, which named SCHOG, to construct a human hand detector. In first, we extract our SCHOG feature by combining HOG with skin color cues. Secondly, we apply support vector machine (SVM) algorithm for training our dataset and construct a SVM trained classifier for hand detection. Finally, we test our method on the testing dataset for the SCHOG features and the unchanged HOG features respectively. As experimental results shown, SCHOG exhibits a good performance on our testing dataset.