Authors : S. Anila and N. Devarajan
Abstract: Visual recognition is a process of authenticating a true identity of a person. It is one of the applications of Biometrics. This technique makes it possible to use a person’s image to verify their identity. The proposed system includes three different modules such as compression, feature extraction and feature matching. The face recognition is performed in the compression domain. Three transforms are utilized for compression namely Discrete Wavelet Transform (DWT), Discrete Cosine Transform (DCT) and Fast Fourier Transform (FFT). Principle Component Analysis (PCA) and Euclidean distance algorithm is used for feature extraction and feature matching. By taking transform the image enters into compression domain and inverse transform is omitted. As a result, computational complexity, time consumption and storage space of the visual recognition system is reduced.
S. Anila and N. Devarajan, 2013. An Efficient Face Recognition System Using Transforms. International Journal of Soft Computing, 8: 42-49.