摘要
提出一种改进的基于连通区域标记的人数检测算法。用实时的图像与教室背景图像作背景差分分割出目标人员;利用人脸肤色模型和头发颜色信息对人员头部进行不同颜色的标记;针对处理之后的三值图像,用改进的连通区域标记算法对人员进行检测,检测结果可对建筑内自动照明系统控制和自习室管理提供依据。实验结果表明,该方法对于检测教室人数具有较高的准确率。
An improved people counting detection algorithm based on connected labeling is proposed. In this method,background subtraction method is used to achieve the segmentation of foreground from background. For human detection,skin and hair are marked in different colors based on skin model and hair color information. Based on the three values image,an improved connected labeling method is used to detect human,the test results can be used in automatic lighting control system and classroom management. Experimental results show that this method has a higher detection rate.
出处
《沈阳理工大学学报》
CAS
2016年第1期17-21,共5页
Journal of Shenyang Ligong University
关键词
目标分割
头部标记
连通区域标记
人员检测
human detection
object segmentation
head labeling
connected labeling