Abstract—In this paper, we propose a fast mouth detection scheme. For mouth detection, there are many related techniques which could not exactly locate the correct mouth position, especially profile face or light influence. In order to solve this problem, we propose an approach to address this problem based on the effective skin-color segmentation filter and edge projection to detect the mouth feature. The main procedures of this approach consist of face extraction, mouth feature extraction, and mouth detection. For face extraction, we use the Viola and Jones detector to extract the face region. In order to more effectively and correctly detect the mouth position, we adopt two phases which are the correct facial features extraction and the projection computation for the specified area in skin region to achieve the purpose. The results show that our approach can obtain a high accuracy and a fast executed time with face extraction and mouth feature localization.
Index Terms—Adaboost algorithm, skin color transformation, Canny detector, projection.
The authors are with the Department of Computer Science and Information Engineering, National Formosa University, Taiwan (phone: +886-5-631-5590; fax: +886-5-633-0456; e-mail: hyhuang@nfu.edu.tw, v77331@gmail.com).
[PDF]
Cite:Hui-Yu Huang and Yan-Ching Lin, "An Efficient Mouth Detection Based on Face Localization and Edge Projection," International Journal of Computer Theory and Engineering vol. 5, no. 3, pp. 514-519, 2013.