Robust smile detection algorithm using curvature and area as cues

Authors

Karl Bryan A. Lagman ⋅ PH National Institute of Physics, University of the Philippines Diliman
Michael Andrei C. Paguio ⋅ PH National Institute of Physics, University of the Philippines Diliman

Abstract

We presented a computer vision approach to classify whether the person is smiling or not. We used the curvature and area of the lips as features in detecting smile. We have detected the smile even if the teeth are not visible using curvature as cue, and, even if the curvature of the lips is not well defined, using area as cue. We have obtained an 87% recognition in classification with area as an added cue.

Downloads

Issue

2006: Proceedings of the 24th Samahang Pisika ng Pilipinas National Physics Congress

Article ID

SPP-2006-PB-04

Section

Poster Session PB

Published

2006-10-25

How to Cite

[1]

KBA Lagman and MAC Paguio, Robust smile detection algorithm using curvature and area as cues, Proceedings of the Samahang Pisika ng Pilipinas 24, SPP-2006-PB-04 (2006). URL: https://proceedings.spp-online.org/article/view/SPP-2006-PB-04.