A Facial Expression Classification using Histogram Based Method

With a number of emerging new applications, automatic recognition of facial expressions is a research area of current interest. In this paper, a simple and effective approach for accurate facial expression recognition using morphological based method along with histogram method had been proposed. This paper focuses on facial expression to identify five universal human expressions: neutral, happy, anger, sad and surprise. Proposed system consists of four stages. In the first stage, using median filter which reduces noise that cause over segmentation and pre-processes the face image. In the second stage, mouth region is segmented using the morphological method. After segmentation part, the geometric feature (especially mouth) is extracted in the third stage. Lastly, histogram calculations are used for facial expression classification. The proposed system tested on JAFFE facial expression database. The effectiveness of the proposed system can be confirmed through the experimental results.