作者: Y. Aidarous , R. Seguier , S. Le Gallou , A. Sattar
DOI:
关键词: Face (geometry) 、 Nelder mead simplex 、 Computer vision 、 Computer science 、 Efficient algorithm 、 Simplex 、 Pose 、 Access time 、 Image (mathematics) 、 Active appearance model 、 Artificial intelligence
摘要: In this paper we propose an efficient algorithm to align the face in real time, based on Active Appearance Model (AAM) 2.5D. The main objective is make a robust, rapid and memory application suitable for embedded systems, so they could pose rapidly by using less memory. Classical AAM high consumer algorithm, consequently transfer of stored system makes it time consuming as well. Our 2.5D generated taking 3D landmarks from frontal profile view 2D texture only image. Moreover pro- Nelder Mead Simplex technique search. It does not require large memory, thus becoming systems elimi- nating excess access requirements. We illustrate optimized estimation test three databases: M2VTS, synthetic images webcam images. Results vali- date our combination simplex