中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊

Journal of Chongqing Jiaotong University(Natural Science) ›› 2013, Vol. 32 ›› Issue (1): 95-98.DOI: 10.3969/j.issn.1674-0696.2013.01.22

Previous Articles     Next Articles

Driver’s Face Positioning Method Based on Color Space and Face

Fu Kai,Ge Ruhai,Fu Hongyu,Jin Qiao   

  1. School of Automotive and Traffic Engineering,Jiangsu University,Zhenjiang 212013,Jiangsu,China
  • Received:2012-01-04 Revised:2012-04-19 Online:2013-02-15 Published:2015-03-12

基于颜色空间和面部特征的驾驶员面部定位方法

符凯,葛如海,符鸿玉,金桥   

  1. 江苏大学汽车与交通工程学院车辆工程系,江苏镇江212013
  • 作者简介:符凯(1987—),男,江苏南通人,硕士研究生,主要从事车辆主动安全、安全辅助驾驶技术方面的研究。E-mail:kevin19871228@ hotmail.com。

Abstract: A fast method for the detection and location of the driver’s facial area was introduced. Firstly the image was mapped into YCbCr space and the skin Gaussian model was built. Secondly driver’s face area was got by likelihood figure adaptive threshold segmentation. The processed binarization image should be carried out morphology operation to eliminate burrs and holes. Finally the facial image was processed with gray integral projection to get driver’s face area. The test shows this method has effectiveness and timeliness.

Key words: face detection, skin model, YCbCr color space, morphology, gray integral projection

摘要: 介绍了一种快速检测和定位驾驶员面部区域的方法。首先将驾驶员面部图像映射到YCbCr 颜色空间,建立 肤色高斯模型; 然后通过似然图自适应阈值分割获得驾驶员面部区域,对获得的二值化面部图像进行形态学操作消 除毛刺和小孔; 最后对处理后的面部图像进行灰度积分投影,结合面部特征获得驾驶员的面部区域并框定。并试验 证明了方法的实时有效性。

关键词: 人脸检测, 肤色模型, YCbCr颜色空间, 形态学, 灰度积分投影

CLC Number: