摘要
为了提高人脸识别的识别率,提出了一种基于自适应对数变换和主成分分析(PCA)算法的人脸识别方法。将人脸图像进行自适应对数变换,使人脸图像由于光照不均而引起的图像模糊得到改善;使用PCA算法对图像进行降维和特征提取,减少了图像识别的计算量,有效提高识别的效率,再用最近邻分类器(NN)进行分类识别。在ORL和Yale人脸数据库上进行了使用验证,结果表明该方法能够提高人脸识别的识别率。
To improve the recognition rate of facial recognition, a novel method based on adaptive logarithmic transformation and Principal Components Analysis （PCA） is presented. Firstly, adaptive logarithmic transformation is carried out to the face images, which improves the image fuzzy caused by illumination variations. Then, the images are processed by PCA algorithm to reduce di- mensionality and extract features, so the calculation amount is reduced and recognition rates are improved. Then, the nearest neighbor classifier is selected as classifier for the recognition of face images. Finally, the ORL and Yale face database is used to test the proposed method. The experiment results show that the method can improve the recognition rate.
出处
《电子技术应用》
北大核心
2014年第6期126-129,共4页
Application of Electronic Technique
基金
国家自然科学基金项目(11202106)
教育部高等学校博士学科点专项科研基金项目(20123228120005)
江苏省高校自然科学研究计划项目(13KJB170016)
关键词
人脸识别
自适应对数变换
主成分分析
最近邻分类器
face recognition
adaptive logarithmic transformation
principal components analysis
nearest neighbor
