Human Facial Expression Recognition Based On Fractional Fourier Transform

Over the last decade, automatic facial expression analysis has become an active research area which finds potential applications in fields such as more engaging human-computer interaction, multimedia information analysis and retrieval, biometrics for security and surveillance, entertainment and e-health having huge crucial researching values. In general, there are three steps in the process of facial expression recognition including:face detection, facial features extraction, classification of emotions. In this paper, it mainly makes researches on the above mentioned steps and puts Fractional Fourier Transform (FrFT) which is a booming tool in the field of time-frequency analysis into recognizing human emotional state. As the generalized form of the Fourier transform, the FrFT can be interpreted as a rotation of signals in the time-frequency plane. The analyzed signals can be mapped to any domain between time domain and frequency domain with the method of FrFT at will to extract images features more effectively. Therefore, the FrFT not only can achieve the functions as the traditional theories but also be more universal and flexible. As a result, the FrFT is hoped to obtain better results in the field of facial expression recognition than the traditional theories. The main work of this paper is organized as follows:1. It describes the significance of FrFT and analyzes the existing various discrete algorithms of FrFT. Moreover, the crucial researched discrete method is provided by Ozaktas. Besides, it also extends the one dimension form of FrFT into the form of two dimensions (2D-FrFT) based on the mathematics and physical meaning and achieves the discrete algorithm of 2D-FrFT. Simultaneously, it also makes some necessary discussions on the properties of 2D-FrFT. 2. It makes illustrations on selecting emotional database and pre-processing of the facial database images in detail. Furthermore, we put the researching emphasizes on explaining the properties of 2D-FrFT in the field of image processing and definite the regulation on how to extract the features from the facial images. At the same time, it also reduces the dimensions of the original images effectively with the method of nearest neighbor interpolation which will improve the character of real-time in the above facial expression recognition process in some degree.3. It introduces the Fisher classification algorithm in general and constructs the model of recognizing human facial expression based on the method of 2D-FrFT. In additional, it also quantitative analyzes the relation between the transform orders of 2D-FrFT and the recognizing results. In the last, it compares the results in this paper and the results with the methods of Gabor to interpret the validity of 2D-FrFT in the field of facial expression recognition.