摘要: 提出了Real-Adaboost的一种改进算法。该算法采用预先计算类Haar特征所对应弱分类器在样本空间的划分,并动态更新人脸训练样本的权值。与以往的Real-Adaboost算法比较,该算法大大缩短了训练时间,算法训练时间复杂度降到O(T*M*N),同时加速了强分类器的收敛性能,减少检测器的弱分类器数量,减少检测时间。
关键词:
人脸检测,
实值Adaboost,
类Haar特征,
层叠分类器,
动态权值
Abstract: This paper proposes a novel human face detection algorithm based on real Adaboost algorithm. Policy that calculates in advance the partitioning of Haar-like feature weak classifiers in sample input space and updating training face samples’ weights dynamically is adopted. This algorithm reduces training time cost greatly compared with classical real-Adaboost algorithm. In addition, it speeds up strong classifier converging, reduces the number of weak classifiers and decreases detecting time.
Key words:
Face detection,
Real-Adaboost,
Haar-like feature,
Cascade classifier,
Dynamic weight
武 妍;项恩宁. 动态权值预划分实值Adaboost人脸检测算法[J]. 计算机工程, 2007, 33(03): 208-209.
WU Yan; XIANG Enning. Dynamic Weights and Pre-partitioning Real-Adaboost Face Detection Algorithm[J]. Computer Engineering, 2007, 33(03): 208-209.