WANG Jinwei,HU Bufa.3D face recognition method based on local multi - feature fusion[J].Journal of Yanbian University,2022,(03):242.
基于局部多特征融合的三维人脸识别方法
- Title:
- 3D face recognition method based on local multi - feature fusion
- 文章编号:
- 1004-4353(2022)03-0242-05
- Keywords:
- 3D face recognition; window; key points; feature vector
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 为了克服表情变化对三维人脸识别的影响,提出一种基于局部多特征融合的三维人脸识别方法.该方法首先根据中心侧影线提取鼻尖点,并以鼻尖点作为基准点制定窗口; 然后利用形状索引值在窗口内提取关键点,并计算每个关键点和其区域的多维度特征后将其融合成特征向量; 最后采用相似度匹配方法进行人脸识别,并以匹配点数最多的特征向量作为最终的识别结果.实验结果表明,该方法的识别率到达97.7%,且具有较好鲁棒性,同时优于文献[4]、[6]和文献[7]的方法; 因此,该方法可为有效解决表情变化对三维人脸识别的影响提供参考.
- Abstract:
- In order to overcome the influence of expression changes on 3D face recognition, a 3D face recognition method based on local multi - feature fusion is proposed.Firstly, the method extracts the nose tip according to the center profile line, and uses the nose tip as the reference point to formulate a window; then uses the shape index value to extract key points in the window, and calculates the multi - dimensional features of each key point and its area to fuse them into a Feature vector.Finally, the similarity matching method is used for face recognition, and the feature vector with the most matching points is used as the final recognition result. The experimental results show that the recognition rate of this method reaches 97.7%, and it has good robustness, which is better than the methods in the literature [4], [6] and [7]. Therefore, this method can provide a reference for effectively solving the influence of expression changes on 3D face recognition.
参考文献/References:
[1] PATIL H, KOTHARI A, BHURCHANDI K.3D face recognition: features, databases, algorithms and challenges[J].Artificial Intelligence Review,2015,44(3):393 - 441.
[2] 罗常伟,於俊,于灵云,等.三维人脸识别研究进展综述[J].清华大学学报(自然科学版),2021,61(1):77 - 88.
[3] HARIRI W, TABIA H, FARAH N, et al.3D face recognition using covariance based descriptors[J].Pattern Recognition Letters, 2016,78:1 - 7.
[4] 郭梦丽,达飞鹏,邓星,等.基于关键点和局部特征的三维人脸识别[J].浙江大学学报(工学版),2017,51(3):584 - 589.
[5] EMAMBAKHSH M, EVANS A.Nasal patches and curves for expression - robust 3D face recognition[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017,39(5):995 - 1007.
[6] LI Y, WANG Y H, LIU J, et al.Expression - insensitive 3D face recognition by the fusion of multiple subject - specific curves[J].Neurocomputing, 2018,275:1295 - 1307.
[7] 张红颖,杨维民,王汇三.结合局部关键点集与测地线的三维人脸识别[J].激光与光电子学进展,2020,57(22):1 - 8.
[8] XU C H, WANG Y H, TAN T N, et al.Combining local features for robust nose location in 3D facial data[J].Pattern Recognition Letters, 2006,27(13):1487 - 1494.
[9] DORAI C, JAIN A.COSMOS: a representation scheme for 3D free - formobjects[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 1997,19(10):1115 - 1130.
[10] QUAN W, MATUSZEWSKI B J, SHARK L K.Statistical shape modelling for expression - invariant face analysis and recognition[J].Pattern Analysis and Applications, 2016,19(3):765 - 781.
[11] HARIRI W, FARAH N, VISHWAKARMAL D K.Deep and shallow covariance feature quantization for 3D facial expression recognition[EB/OL].(2021-05-12)[2022-02-03].https://arxiv.org/abs/2105.05708.
备注/Memo
收稿日期: 2022-04-14
基金项目: 黎明职业大学校课题(LZ2019113); 福建省中青年教师教育科研项目(JZ180904)
作者简介: 王金伟(1986—),男,硕士,讲师,研究方向为机器视觉、模式识别.