摘要
Boosting是一种有效的分类器组合方法,它用某个分类算法生成一系列的基分类器,每个基分类器的训练依赖于在其之前产生的分类器的分类结果,基分类器在训练集上的错误率用于调整训练样本的概率分布,最终分类器通过单个基分类器的加权投票建立起来。将Boosting算法应用在动态车型图像检测中,大大提高了对运动过程中车辆的识别能力,对智能交通系统的发展起着推动作用。
Boosting is an effective method of classifier combination, which uses a classifier algorithm to generate a series of base classifiers, each basing its training on the classified result generated before it, while error rate of base classifier on Training Set is used to adjust the probability distribution of training samples, and the final classifier is established by the weighted vote of single base classifiers. In this paper, boosting is used in dynamic algorithm model image detection, greatly improving the recognizability of vehicles in motion, and promoting the development of intelligent transportation systems.
出处
《河北工业科技》
CAS
2008年第5期310-311,338,共3页
Hebei Journal of Industrial Science and Technology
基金
渭南师范学院科研基金资助项目(08YKS027)
