Facial recognition based on straight angle fuzzy texture unit matrix

Efficient face recognition is one of the long standing problems of computer vision. Texture is an important spatial feature, useful for identifying facial objects of interest in an image. The most popular statistical methods used to measure the textural information of images are the grey level cooccurrence matrix (GLCM) and the texture spectrum (TS) Approach. The major problem of these two approaches is the dimensionality. This paper derives significant local information by deriving a new integrated matrix using fuzzy based texture unit and GLCM for efficient face recognition. The features derived from this integrated matrix provide detailed and complete texture information of about the human faces. Local binary pattern (LBP) operator is a powerful operator to extract the discriminative facial features and this method may fail to detect the illumination variation and facial expressions accurately. Noise is the crucial problem related to methods based on LBP also. To address these issues, the present paper derives a fuzzy logic on texture unit to discriminate the local information precisely than LBP and to reduce the dimensionality. The performance of the proposed scheme is validated using complex facial datasets, namely Yale and american telephone and telegraph company (AT&T). The present method is compared with methods based on LBP and the results indicate the efficacy of the proposed method.