• Home
  • Current congress
  • Public Website
  • My papers
  • root
  • browse
  • IAC-20
  • B1
  • VP
  • paper
  • electricity towers detection and mapping using khalifsat satellite images with deep learning techniques

    Paper number

    IAC-20,B1,IP,11,x57662

    Author

    Ms. Alya AlMaazmi, United Arab Emirates, Mohammed Bin Rashid Space Centre (MBRSC)

    Coauthor

    Mr. Saeed Al Mansoori, United Arab Emirates, Mohammed Bin Rashid Space Centre (MBRSC)

    Coauthor

    Ms. Meera AlShamsi, United Arab Emirates, Mohammed Bin Rashid Space Centre (MBRSC)

    Year

    2020

    Abstract
    Electricity is one of the major power source worldwide. Some countries such as the developing nations still have limited access to electricity. These nations lack a complete and accurate map of existing High-Voltage infrastructure, resulting challenges in improving the electric grid. High resolution satellite imagery and machine learning algorithms can aid to detect and count the electricity towers for accurate and better mapping.
    Deep Darknet-53 with 53 convolution network has been used to extract electricity towers features from Khalifasat imagery including average pool connected and softmax layers to predict the bounding box, objectness and class predictions.
    The used deep learning network showed promising results on khalifa-sat images, where most of the electricity towers detected with accurate bounding boxes.
    Abstract document

    IAC-20,B1,IP,11,x57662.brief.pdf

    Manuscript document

    (absent)