Enter Note Done
Go to previous page in this tab
Session
  • Presentation | IN44A: Advancing Artificial Intelligence for Remote Sensing: Overcoming Data Scarcity and Domain Shift II Oral
  • Oral
  • Bookmark Icon
  • IN44A-02: Developing a Neighborhood-Relation Poverty Mapping Method Using Household Survey and Satellite Imagery
  • Schedule
    Notes
  • Location Icon293
    NOLA CC
    Set Timezone
  •  
    View Map

Generic 'disconnected' Message
Author(s):
Tunyu Chang, University of Tokyo (First Author, Presenting Author)
Akiyuki Kawasaki, The University of Tokyo


The Lack of poverty data in developing countries makes satellite-based poverty mapping a valuable tool for estimating economic conditions. Household-level poverty prediction using directly collected data can reduce errors caused by anonymity compared to other larger scale poverty mapping method. Previous studies have constructed a model to predict household poverty in rural areas in developing countries. However, the applicability of such a model in suburban areas remains underexplored. We propose a new household-level poverty mapping approach that is novel in using satellite and survey data. By considering temporal rooftop changes and neighborhood clustering, we generate a fine-resolution poverty map validated by the HIS data from Marikina, Philippines. Our method enables scalable accurate poverty mapping and provides a quantitative approach to assess poverty over large areas. Such detailed poverty maps can enhance the efficiency of humanitarian aid distribution and support studies on the spatial dynamics of poverty, such as vulnerability to natural disasters like flooding. By offering finer spatial targeting, this approach could help researchers and policymakers better identify and respond to localized needs.



Scientific Discipline
Suggested Itineraries
Neighborhood
Type
Main Session
Discussion