欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
人脸识别是图像处理领域的一个热点。由于红外人脸识别可以避免可见光人脸识别存在的一些固有的缺陷，因此有着广阔的应用前景。文中从统计学角度和生物特征角度提出基于贝叶斯分类和血流模型的红外人脸识别方法，这种方法可以充分利用人脸血流模型的优势，减弱环境因素对红外人脸识别的性能的影响，提取精确的生物学特征，同时根据统计特征，并使用贝叶斯分类器，增加样本之间的类间距，减少样本之间的类内距。该方法将人脸温谱图转换为人脸血流图；使用PCA算法对人脸血流模型数据进行降维处理，并训练产生内部子空间和外部子空间；通过贝叶斯分类算法进行人脸识别。文中按照这个思路做了对比实验，实验结果证明这种方法是行之有效的。
英文摘要：
Face recognition is a hot research topic in the field of image processing. Infrared face recognition avoids some inherent defects of visible light face recognition,therefore has a broad application prospects. From biological characteristics and a statistical point, an infrared face recognition method based on face blood perfusion model and Bayesian classification is proposed in this paper, which takes full advantage of the blood perfusion model of the human face, weakens the impact of environmental factors on the performance of infrared face recognition and extracts the precise biological characteristics. At the same time, the method enlarges between-class distance and less- ens within-class distance according to the statistical characteristics. Thermal images are converted into blood perfusion data;the PCA method is used to reduce the dimension of the face blood perfusion model data and generate the internal subspace and external subspac~; the Bayesian classifier is used for face recognition. According to this idea,the comparative experiments are made in this paper,and this method is proved effective by the results of the experiments.
同期刊论文项目
同项目期刊论文
位置：