Residential scene classification for gridded population sampling in developing countries using deep convolutional neural networks on satellite imagery.

Journal: International journal of health geographics
PMID:

Abstract

BACKGROUND: Conducting surveys in low- and middle-income countries is often challenging because many areas lack a complete sampling frame, have outdated census information, or have limited data available for designing and selecting a representative sample. Geosampling is a probability-based, gridded population sampling method that addresses some of these issues by using geographic information system (GIS) tools to create logistically manageable area units for sampling. GIS grid cells are overlaid to partition a country's existing administrative boundaries into area units that vary in size from 50 m × 50 m to 150 m × 150 m. To avoid sending interviewers to unoccupied areas, researchers manually classify grid cells as "residential" or "nonresidential" through visual inspection of aerial images. "Nonresidential" units are then excluded from sampling and data collection. This process of manually classifying sampling units has drawbacks since it is labor intensive, prone to human error, and creates the need for simplifying assumptions during calculation of design-based sampling weights. In this paper, we discuss the development of a deep learning classification model to predict whether aerial images are residential or nonresidential, thus reducing manual labor and eliminating the need for simplifying assumptions.

Authors

  • Robert F Chew
    Center for Data Science, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA. rchew@rti.org.
  • Safaa Amer
    Division for Statistical and Data Sciences, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.
  • Kasey Jones
    Center for Data Science, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.
  • Jennifer Unangst
    Division for Statistical and Data Sciences, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.
  • James Cajka
    Geospatial Science and Technology Program, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.
  • Justine Allpress
    Geospatial Science and Technology Program, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.
  • Mark Bruhn
    Geospatial Science and Technology Program, RTI International, 3040 East Cornwallis Road, Research Triangle Park, NC, USA.