New Method for Classification of Age Groups Based on Texture Shape Features

Abstract

Recently Automatic age classification based on human faces has attained an increasing attention by many researchers due to it’s wide spread applications. One of the problems with the recent age classification and recognition approaches are that they have to extract features on the entire image and on the large grey level range of the image. To overcome this problem present paper derived Texture Shape Features(TSF) using fuzzy logic on Image Dimensionality Reduction using Shape Primitives(IDRSP) model for an efficient age classification. The proposed Fuzzy IDRSP(FIDRSP) model reduces the dimensionality of the image using shape primitives and reduces the grey level range by using a fuzzy logic while preserving the significant attributes of the texture. The present paper derived TSF on the FIDRSP model on only two parts of the facial skin namely chin and forehead instead of entire face and found that these features drastically affect the age classification system of humans. The present paper using TSF’s on FIDRSP model attempted the tedious and complex task of classification of human age into five different groups i.e. 0-12, 13-26, 27-45, 46-60 and above 60. The TSF’s are derived from the patterns formed by Bezier curve (BC), and U, V and T patterns on the facial skin. The patterns are measured on a 5×5 mask, because most of the curve and complex shape primitives properties do not fit into 3×3 mask. The experimental results indicate the efficacy of the proposed method over the other methods.