Victim Localization in Urban Search and Rescue using Multi-layer Maps

  • Abdulrahman Goian

Student thesis: Master's Thesis


Abdulrahman S. Goian. 'Victim Localization in Urban Search and Rescue using Multi-layer Maps', M.Sc. Thesis, M. Sc. by Research in Engineering, Department of Electrical and Computer Engineering, Khalifa University of Science and Technology, United Arab Emirates, February 2018. Urban search and rescue missions are critical in the sense that they require quick intervention to locate victims and survivors before their state become worse. In an attempt to facilitate this activity, Unmanned Arial Vehicles (UAV) are used to examine the disaster scene in search of possible victims. The vehicle is equipped with multiple sensors allowing it to generate multiple reference maps. Advanced deep learning, wireless based localization, and thermal mapping techniques are used to incrementally build victim location maps. These maps can then be used either separately or as a combined map to navigate the UAV using an iterative process of selecting the local best place to go to in order to maximize the chance of locating victims. In this thesis, we propose NBV (next best view) approach that utilizes a multi-objective utility function to guide the UAV, using the multi-sensory maps and an adaptive sampling approach. Simulation results demonstrate that the proposed multi-objective utility function outperforms commonly used utilities in many several key performance measures such as less number of iterations, lower entropy reduction and shorter traveled distance. In addition, implementing the proposed utility with the multi-sensory maps yields better performance when compared to the individual map performances tested using the proposed utility or other commonly used utilities.
Date of AwardFeb 2018
Original languageAmerican English
SupervisorNawaf Al Moosa (Supervisor)


  • Victim Localization
  • Multi-layer Maps
  • Next Best View
  • Information Gain.

Cite this