• Submitted by: Ali Ghali
    Supervised by Dr Bassam Kurdy

    Journal of Theoretical and Applied Information Technology

    ISSN: 1992-8645

    E-ISSN: 1817-3195

    Date

    30th September 2018. Vol.96. No 18

    ABSTRACT

    Emotion recognition is a natural capability in human beings. However, if we are to ever create a humanoid “robot” that can interact and emote with its human companions, the difficult task of emotion recognition will have to be solved. The ability for a computer to recognize human emotion has many highly valuable real world applications. Consider the domain of therapy robots which are designed to provide care and comfort for infirm and disabled individuals. These machines could lever information on a patient’s current and evolving state of mind, in order to tailor personalized strategies for patient care and interaction. For example, when a patient is upset or unhappy, a more effective strategy may be take a moment to recognize the emotion and offer sympathy. Even outside of the realm of robotics, working with computers that have the ability to sense and respond to emotional state can go a long way to improve the quality of human-computer interaction (HCI). By designing HCI to be more like human-human interaction, we have the ability to create more natural, fulfilled, and productive working relationships with our machines. In this research we explain how to recognize emotions through digital images using Android application, and we will identify seven types of emotions (neutral- happy- sad- surprised- afraid- angry- disgusted). We designed this work based on a popular library called OpenCv, and the Fisherfaces algorithm that consists of (PCA) principle component analysis algorithm and (LDA) the linear discriminate analysis algorithm, in addition, we built the server using Java language to implement the android application, also we compare the coordinates of eyes and mouth in test image with the coordinates in the database to take the highest similarity and show the result. The language used to build this work is the Java language using NetBeans IDE 8.0.2, and the use of android studio to design android application.

    http://www.jatit.org/volumes/Vol96No18/11Vol96No18.pdf

Comparison Study Of Automatic Classifiers Performance In Emotion Recognition Of Arabic Social Media Users