ECG Identification Based on PCA-RPROP

  • Jinrun Yu
  • Yujuan Si
  • Xin Liu
  • Dunwei Wen
  • Tengfei Luo
  • Liuqi Lang
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10287)


With the quick development of information technology, people pay more and more attention to information security and property safety, where identity is one of the most important aspects of information security. Compared with the traditional means of identification, biometrics recognition technology offers greater security and convenience. Among which, electrocardiogram (ECG) human identification has been attracted great attention in recent years. As a new type of biometric feature authentication technology, the feature selection and classification of ECG has become a focus of the research community. However, there exist some problems that can impair the efficiency and accuracy of ECG identification, including information redundancy and high dimensionality in feature extraction, and insufficient stability in classification. In order to solve the problems, in this paper, we propose a recognition method based on PCA-RPROP. In this method, firstly, only R points are located to get the original single-cycle waveforms. Then, PCA and whitening are used to process original data, where whitening is to make the input less redundant and PCA is to reduce its dimensionality. Finally, the resilient propagation (RPROP) algorithm is used to optimize the neural network and establish a complete recognition model. In order to evaluate the effectiveness of the algorithm, we compared the PCA feature with the wavelet decomposition and multi-point localization features in an ECG-ID database, and also compared RPROP with traditional BP algorithm, SVM and KNN. The experimental results show that this method can improve the performance compared with other classifiers, and simultaneously reduce the complexity of localization and the redundancy of features. It is superior to the other methods both speed and accuracy in recognition, especially when compared with the traditional BP. It can solve the problems of traditional BP with 2.4% higher recognition accuracy than LIBSVM, and 14 s faster than KNN in terms of time efficiency. Therefore, it is an efficient, simple and practical recognition algorithm.


ECG Identity recognition Whitening PCA dimensionality reduction Neural network RPROP 



We thank all the volunteers and colleagues provided helpful comments on previous versions of the manuscript. This work was supported by the Science and Technology Development projects funded by Jilin Government (20150204039GX, 20170414017GH), Science and Technology Development Special Funded by Guangdong Government (2016A030313658), and Key Scientific and Technological Special Fund Project supported by Changchun Government under Grant No. 14KG064. This work was also supported by Premier-Discipline Enhancement Scheme Supported by Zhuhai Government and Premier Key-Discipline Enhancement Scheme Supported by Guangdong Government Funds.


  1. 1.
    Biel, L., Pettersson, O., Philipson, L., et al.: ECG analysis: a new approach in human identification. IEEE Trans. Instrum. Meas. 50, 808–812 (2001)CrossRefGoogle Scholar
  2. 2.
    Palaniappan, R., Krishnan, S.M.: Identifying individuals using ECG beats. In: International Conference on Signal Processing and Communications, Bangalore, India, pp. 569–572 (2005)Google Scholar
  3. 3.
    Israel, S.A., Irvine, J.M., Cheng, A., et al.: ECG to identify individuals. Pattern Recogn. 38, 133–142 (2005)CrossRefGoogle Scholar
  4. 4.
    Plataniotis, K.N., Hatzinakos, D., Lee, J.K.M.: ECG biometric recognition without fiducial detection. In: Biometric Consortium Conference, 2006 Biometrics Symposium: Special Session on Research, Baltimore, MD, pp. 1–6 (2006)Google Scholar
  5. 5.
    Jégou, H., Chum, O.: Negative evidences and co-occurrences in image retrieval: the benefit of PCA and whitening. In: Fitzgibbon, A., Lazebnik, S., Perona, P., Sato, Y., Schmid, C. (eds.) ECCV 2012. LNCS, pp. 774–787. Springer, Heidelberg (2012). doi: 10.1007/978-3-642-33709-3_55 CrossRefGoogle Scholar
  6. 6.
    Moré, J.J.: The Levenberg-Marquardt algorithm: implementation and theory. In: Watson, G.A. (ed.) Numerical Analysis. LNM, vol. 630, pp. 105–116. Springer, Heidelberg (1978). doi: 10.1007/BFb0067700 CrossRefGoogle Scholar
  7. 7.
    Riedmiller, M., Braun, H.: A direct adaptive method for faster backpropagation learning: the RPROP algorithm. In: IEEE International Conference on Neural Networks, San Francisco, CA, vol. 1, pp. 586–591 (1993)Google Scholar
  8. 8.
    Saxena, S.C., Kumar, V., Hamde, S.T.: Feature extraction from ECG signals using wavelet transforms for disease diagnostics. Int. J. Syst. Sci. 33, 1073–1085 (2002)CrossRefzbMATHGoogle Scholar
  9. 9.
    Han, X.U., Wang, D.D., Jiang, T.B.: Research on the ECG signal denoising algorithm based on wavelet transform and the median filter. Autom. Instrum. (2012)Google Scholar
  10. 10.
    Sadhukhan, D., Mitra, M.: R-peak detection algorithm for ECG using double difference and RR interval processing. Procedia Technol. 4, 873–877 (2012)CrossRefGoogle Scholar
  11. 11.
    Marrtis, R.J., Acharya, U.R., Min, L.C.: ECG beat classification using PCA, LDA, ICA and discrete wavelet transform. Biomed. Signal Process. Control 8, 437–448 (2013)CrossRefGoogle Scholar
  12. 12.
    Chang, C.C., Lin, C.J.: LIBSVM: a library for support vector machines. ACM Trans. Intell. Syst. Technol. 2, 27 (2011)CrossRefGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Jinrun Yu
    • 1
  • Yujuan Si
    • 1
    • 2
  • Xin Liu
    • 1
  • Dunwei Wen
    • 3
  • Tengfei Luo
    • 1
  • Liuqi Lang
    • 2
  1. 1.College of Communication EngineeringJilin UniversityChangchunChina
  2. 2.Zhuhai College of Jilin UniversityZhuhaiChina
  3. 3.School of Computing and Information SystemsAthabasca UniversityAlbertaCanada

Personalised recommendations