ELASTIC FACE RECOGNIZER: INVARIANT FACE RECOGNITION BASED ON ELASTIC GRAPH MATCHING MODEL
Raymond Lee · International Journal of Pattern Recognition and Artificial Intelligence · 2002
Human faces are difficult to interpret because they are highly variable. Over the last few decades, various techniques have been proposed for computer recognition of human faces. In this paper, we introduce an Elastic Graph Dynamic Link Model to automate the process of facial recognition. It is integrated with the Active Contour Model to provide an effective and efficient means of facial contour extraction and recognition. A portrait gallery of 100 distinct facial images is used for network training. Experimental results are presented for a database of 1,020 tested face images, which were obtained under conditions of widely varying facial expressions, viewing perspectives and image sizes. An overall average correct recognition rate of over 86% is attained.