Image Compression in Wireless Sensor Networks Using Autoencoder and RBM Method

  • S. Aruna DeepthiEmail author
  • E. Sreenivasa Rao
  • M. N. Giri Prasad
Conference paper
Part of the Lecture Notes in Networks and Systems book series (LNNS, volume 65)


Wireless sensor networks are used in various day-to-day applications. With advanced technology, wireless sensor networks (WSNs) play a key role in networking technologies since it can be expanded without communication infrastructure. There is so much demand for wireless sensor networks that are useful in day-to-day applications. Various techniques like multilayer restricted Boltzmann machine (RBM) network and variational autoencoder (VAE) methods are designed for transmission of images in wireless sensor networks and tested for both gray (2D) and color (3D) images in this paper. The proposed variational autoencoder method is compared with multilayer RBM when the layers are 2, 4, and 8, and the peak signal-to-noise ratio (PSNR) and signal-to-noise ratio (SNR) are compared.


RBM Autoencoder VAE WSN 


  1. 1.
    Deepthi SA, Rao ES, Prasad MNG (2018) Image transmission and compression techniques using SPIHT and EZW in WSN. In: 2018 2nd international conference on inventive systems and control (ICISC), Coimbatore, pp 1146–149.
  2. 2.
    Bengio Y, Courville AC, Vincent P (2012) Unsupervised feature learning and deep learning: a review and new perspectives. Department of Computer Science and Operations Research, University of Montreal, Montreal, CanadaGoogle Scholar
  3. 3.
    Cheng C, Wang S, Chen X, Yang Y (2016) A multilayer improved RBM network based image compression method in wireless sensor networks. Int J Distrib Sens Netw (Hindawi Publishing Corporation)Google Scholar
  4. 4.
    Kim J, Kyung C-M (2010) A lossless embedded compression using significant bit truncation for hd video coding. Circuits Syst Video Technol IEEE Trans 20(6):848–860CrossRefGoogle Scholar
  5. 5.
    Vo DT, Solé J, Yin P, Gomila C, Nguyen TQ (2010) Selective data pruning-based compression using high-order edge-directed interpolation. Image Process IEEE Trans 19(2):399–409MathSciNetCrossRefGoogle Scholar
  6. 6.
    Dipayan (2017) Deep learning with Hadoop. Packt PublishingGoogle Scholar
  7. 7.
    Larsen ABL (2016) Autoencoding beyond pixels using a learned similarity metricGoogle Scholar
  8. 8.
    Frans K (2010) Variational autoencoders explained, vol 9, pp 145–152; Desjardins AC, Bengio Y, Parallel tempering for training of restricted Boltzmann machines. J Mach Learn Res Workshop Conf ProcGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  • S. Aruna Deepthi
    • 1
    Email author
  • E. Sreenivasa Rao
    • 2
  • M. N. Giri Prasad
    • 3
  1. 1.Jawaharlal Nehru Technological University AnanthapurAnanthapuramuIndia
  2. 2.Vasavi College of EngineeringHyderabadIndia
  3. 3.Department of ECEJNTU College of Engineering, Jawaharlal Nehru Technological University AnanthapurAnanthapuramuIndia

Personalised recommendations