English

应用科学学报 ›› 2013, Vol. 31 ›› Issue (5): 519-525.doi: 10.3969/j.issn.0255-8297.2013.05.012

• 信号与信息处理 • 上一篇    下一篇

边缘保持滤波与视觉最优准则的人脸图像光照归一化

张二虎， 牟永强， 陈万军   

  1. 西安理工大学信息科学系，西安710048
  • 收稿日期:2012-05-08 修回日期:2013-01-05 出版日期:2013-09-26 发布日期:2013-01-05
  • 作者简介:张二虎，教授，博导，研究方向：图像处理、模式识别与智能信息处理，E-mail: zhangehd@sohu.com
  • 基金资助:

    陕西省重大科技创新项目基金(No.2009ZKC02-17)；西安市科技计划项目基金(No.CXY1127(4))资助

Normalization of Face Illumination Based on Edge-Preserving Filter and Visually Optimal Criterion

ZHANG Er-hu, MOU Yong-qiang, CHEN Wan-jun   

  1. Department of Information Science, Xi’an University of Technology, Xi’an 710048, China
  • Received:2012-05-08 Revised:2013-01-05 Online:2013-09-26 Published:2013-01-05

PDF

1413

摘要： 受计算摄影学中边缘保持滤波器的启发，提出一种人脸图像光照归一化方法，能有效处理复杂光照条件下的人脸识别问题. 该方法采用加权最小均方边缘保持滤波器将亮度图像精确地分解为反射层与阴影层，根据视觉最优准则进行直方图匹配映射，获得视觉质量最优的光照归一化人脸图像. 对YALE-B 与CMU-PIE 人脸数据
库的识别结果表明该方法有效.

关键词: 人脸识别, 光照归一化, 加权最小均方滤波, 视觉最优

Abstract: Inspired by the edge-preserving filter used in computational photography, we propose a method of illumination normalization for face recognition under varying lighting conditions. The brightness layer of a face image is accurately decomposed into a reflectance layer and a shading layer with a weighted least square filter. The histogram of the reflectance layer is remapped to that of sample images selected based on a visually optimal criterion to obtain an image with the best visual quality. Experimental results on the YALE-B and CMU-PIE face databases demonstrate effectiveness of the proposed method under varying illumination conditions.

Key words: face recognition, illumination normalization, weighed least square filter, visually optimal

中图分类号: