基于自适应粒子滤波的ASM新技术

Adaptive Particle Filtering for Facial Feature Tracking

ASM已经被广泛应用于视频流中的人脸跟踪,但大多局限于头部运动较缓慢的正面人脸跟踪,很难精确跟踪复杂背景、恶劣光照和部分遮挡条件下的人脸。本文结合粒子滤波预测跟踪人脸,改进传统ASM方法,实现人脸特征点的精确定位。实验表明,该方法能有效提高ASM搜索的精确度和鲁棒性。

Active shape model (ASM) has been widely used to track a face from a video sequence. However, it is usually limited to frontal view or the cases of small-scale head movement. ASM may fail in condition of cluttered background, adverse circumstances and partial occlusion. An enhanced ASM is proposed to meet those challenges based on particle filtering tracking algorithm. Experiments demonstrate the effectiveness and flexibility of the proposed algorithm in tracking the target objects undergoing significant variation of the object’s appearance or surrounding illumination.