Face tracking system in a dynamic environment
スポンサーリンク
概要
- 論文の詳細を見る
In this paper, we describe a stereovision system able to track faces in a dynamic environment. Pose, orientation, zoom and focus of the controlled cameras are calibrated by an easy and quick method. The face detection is based on the algorithm of Viola et al. A method of validation of face candidates is developed by using epipolar constraint implied by the stereovision and using a determination of the real size of detected objects. The detected face is located in three dimensions in order to be closely followed with updates of pose and zoom of a camera.
- 社団法人電子情報通信学会の論文
- 2003-09-02
著者
-
Kinoshita Keisuke
Atr Human Information Science Labs.
-
Kinoshita Keisuke
Atr Human Information Processing Research Laboratories
-
Yano Sumio
Atr Human Information Science Laboratories
-
Seux Julien
Atr Human Information Science Labs.
-
Yano Sumio
Atr Human Information Science Labs.
関連論文
- Face tracking system in a dynamic environment
- Mouth Tracking from Video Sequences using Trainable Multivariate Gaussian Classifiers
- Mouth Tracking from Video Sequences using Trainable Multivariate Gaussian Classifiers
- 3D Accuracy Improvement Using an Uncalibrated Image