Geospatial Object Detection Using Deep Networks

Barut O., Alatan A. A.

Conference on Earth Observing Systems XXIV, California, United States Of America, 11 - 15 August 2019, vol.11127 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 11127
  • Doi Number: 10.1117/12.2530027
  • City: California
  • Country: United States Of America
  • Keywords: Convolutional Neural Network, Remote Sensing, Object Detection, YOLO, Multiband Satellite Images


In the last decade, deep learning has been drawing a huge interest due to the developments in the computational hardware and novel machine learning techniques. This progress also significantly effects satellite image analysis for various objectives, such as disaster and crisis management, forest cover, road mapping, city planning and even military purposes. For all these applications, detection of geospatial objects has crucial importance and some recent object detection techniques are still unexplored to be applied for satellite imagery. In this study, aircraft, building, and ship detection in 4-band remote sensing images by using convolutional neural networks based on popular YOLO network is examined and the accuracy comparison between 4-band and 3-band images are tested. Based on simulation results, it can be concluded that state-of-the-art object detectors can be utilized for geospatial objection detection purposes.