Smile Detection Techniques in Real-time Video Environment

Abstract

Humans have very good visual system which helps to identify people and objects, perform operations, drive vehicle, read text and so on. Although it might be seen that humans do not take special effort to do most of these tasks. Human visual system is fairly complex to replicate and implement. Computer vision in the simplest term is an automation of human visual system, so that computer and machines obtain high level understanding from digital images and videos. It includes methods for gaining,
processing, examining and understanding digital images and removal of high-dimensional data from the real world to produce numerical or symbolic information. Face detection is the most worthy topic in computer vision and pattern recognition. The mental state of a person is understood by detecting a smile. The proposed system includes capturing a selfie by recognizing smile based on mouth coordinates. It makes use of two different techniques independently, i.e., Dlib Library and Haar cascade classifier for smile detection.