Robust Object Detection Based on Decision Trees and a New Cascade Architecture
Zhiquan Qi, Laisheng Wang, Yitian Xu, Ping Zhong · 2008
In this paper, we describe a robust object detectionmethod using Decision Trees and a new cascade architecture.On the one hand, we design a weak classifier formulti-valued features on AdaBoost algorithm based on DecisionTrees Method, which directly reduces training timeand increases the object detection’s precision. On the otherhand, the use of new cascade architecture is great helpfulfor the problem of minimizing the false accept rateand a cascade’s classification complexity. Experiments onMIT+CMU frontal face data sets and PETS2001 data setsshow that the proposed method is comparable to other existingobject detection method and outperforms the objectdetection method proposed by Ivan Laptev