摘要
为了解决静态背景下的人数识别问题,提出了基于贝叶斯分类的人数识别算法。首先选取固定场景的目标区域,再通过提取该区域图像的HOG特征,进一步训练贝叶斯分类器判别区域是否有人。通过与传统SVM分类效果对比,结果表明该算法能够较好地实现人数识别和人员位置确定。
In order to solve the problem of people counting in static background,a people counting algorithm based on Bayesian classification is proposed.First,the target area of the fixed scene is selected,and then the Bayesian classifier is further trained to determine whether there are people in the area by extracting the HOG feature of the area image.Compared with the traditional SVM classification effect,the results show that the algorithm can better realize the number identification and personnel location determination.
作者
周正宽
田小平
贺玲玲
任继伟
ZHOU Zhengkuan;TIAN Xiaoping;HE Lingling;REN Jiwei(College of Information Science and Technology,Beijing University of Chemical Technology,Beijing 100029,China;Department of Communication Engineering,Beijing Institute of Petrochemical Technology,Beijing 102617,China)
出处
《北京石油化工学院学报》
2019年第4期47-53,共7页
Journal of Beijing Institute of Petrochemical Technology
基金
北京市教委基金资助项目(KM201510017007)
北京市大学生研究训练项目(2017X00008,2018J00230)
关键词
人数识别
贝叶斯分类器
图像分类
特征提取
people counting
Bayesian classifier
image classification
feature extraction
