Year: 2013
Type: article
Abstract: Interest and examining activities in habitual face recognition have increased drastically over the past few years. Faces represent composite, multi-dimensional, significant visual motivation and mounting a computational model for face recognition. For most of the face recognition techniques, solution depends on the feature extraction representation and matching. These lessons are summarized by reflecting the facial expression recognition in general and typically, lack in providing the particular aspect with minimal cost. This, in turn, developed a technique named Color Component Feature Identification using the Bayes Classifier. The model is associated with RGB and HSV color bands along with its corresponding facial feature components. Performance of Color Component Feature Identification using the Bayesian Classifier (CCFI-BC) technique reliably segments the facial color depending on the texture and identifies the features. These regions are further combined with RGB and HSV bands for robust pixel detection and with better visibility. CCFI-BC improves the performance measure and evaluated in terms of recognition rate and true positive rate. A systematic and experiential result shows a minimal cost in restricting the participant’s choice of classifiers.
Authors: Shyla, E. Mary, Punithavalli, M.
Institutions: Bharathiar University, Anna University, Chennai
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7217584560
