Opened or Closed Mouth State Detection for Frontal Face Images by Analysis of using Entropy

Abstract

Specifying the state of opened or closed mouth is a considerable attribute in the international face image standard. This paper presents an algorithm to analyse opened or closed mouth image. The operation is split into two parts; searching system and analysis system. After extraction of the eyes and mouth positions, the mouth image will perform with smoothing, erosion, and Lipmap. The transformed mouth image is calculated the entropy to decide that the mouth is opened or closed. This approach also could give benefit to other applications such as face recognition, travel document evaluation, automatic normalize portrait image, and video transcribe.