A four level aggregative weight scanned model facial quiz image identification

Known Face identification quiz is one of common brain and knowledge mapped quiz. This quiz is defined under different complexities levels for different age groups. These facial quiz having the variation in terms of partial face visibility, scattered face, maze hidden face, highly blurred face etc. In this paper, six of the complex facial quiz problems are been resolved using intelligent level driven analytical model. In this work, four level feature fetch model is presented for analyzing the facial segments. The facial scanning is performed horizontally, vertically and diagonally. A separate window size and shape scan is applied on the facial skin region. The region analysis is done under structured, intensity and feature curve parameters. After generating the weights on each scan, an aggregative measure is assigned to represent the whole image. In second stage, the facial mask is also generated from the quiz image and mapped on each instance of the DB image. The four scanned method is also applied on the DB image under mask implementation. In final stage, the individual scan and aggregative scanned weights are mapped over the DB facial instances and the maximum mapped facial image is identified. The work is applied on randomly collected Indian celebrities facial images and MSRA-CFW celebrities facial database. A larger sample set of 50 and 100 images is tested for six different quiz problems. Four of these complex quiz faces provided the recognition rate over 85% and rest two provided the recognition rate over 90%.