摘要
复杂光照下的人脸识别是模式识别领域一个具有挑战性的问题。通过光照估计提取光照不变量是解决该问题的一种有效方法。在研究朗伯图像获取模型的基础上,提出一种有效的光照估计模型。该模型能够从复杂光照图像中更准确地估计光照,提取光照不变量。Yale B^+复杂光照人脸库的实验结果表明所提算法能够提取更为鲁棒的光照不变量,识别性能优于当前的先进方法。
Face recognition under complex illumination is a challenging task in the field of pattern recognition. Extracting illumina- tion invariants through illumination estimation is an effective way of solving this issue. Based on the research of Lambertian reflec- tance model, an effective illumination estimation model is proposed in this paper. This model can estimate the illumination and ex- tract illumination invariants more accurately from the complex illumination images. Experimental results on the Yale B+ database show that the proposed method extracts more robust illumination invariants and achieves higher recognition performance, compared with the-state-of-the-art methods.
出处
《电视技术》
北大核心
2017年第1期79-83,共5页
Video Engineering
基金
国家自然科学基金青年基金项目(61305011)
江苏省自然科学基金面上项目(BK20131342)
关键词
光照估计模型
光照不变量
朗伯反射模型
人脸识别
illumination estimation model
illumination invariants
Lambertian reflectance model
face recognition