Micro-Expression Recognition Using Histogram of Image Gradient Orientation on Diagonal Planes

Micro-expression conveys a person's state-of-mind; therefore, it can be helpful in many cases. However, its micro-momentary and subtle characteristics make its recognition hard. This challenging task is too arduous in the conditions, where illumination changes. Thus, it is better to use the methods, which are insensitive to shadows and illumination change. Hence, in this paper, we propose the histogram of image gradient orientation on the diagonal planes for the feature extraction. Then, the obtained histogram is fed to the linear support vector machine for the micro-expression recognition. According to the results, the micro-expression recognition rate has been improved using our proposed method in comparison with the state-of-the-art methods.