Murray State Theses and Dissertations

Abstract

Saudi Arabia is a wealthy country with its many resources, but it has seen an increase in poverty recently because of a high rate of population growth with a high rate of unemployment. Some estimate that the number of Saudi Arabians living in poverty is between two and four million. This research aims to develop a way to detect poverty through remote sensing. The study area is Buraydah City, the largest city of the Qassim region, an important agricultural center that plays a significant role in the economy of Saudi Arabia. The research hypothesized that there are poor areas within Buraydah City and aimed to identify them through satellite imagery by using a Landsat 8 Operational Land Imager (OLI) and a high-resolution imagery from the French Système Pour l'Observation de la Terre (SPOT-6) both of 2019. Object-oriented segmentation and classification tools in a Geographic Information System (GIS) were used to distinguish features between poor and richer areas. Three classifiers were applied on the images, which were Maximum Likelihood (ML), Random Tree (RT), and Support Vector Machine (SVM). The best classifier was SVM on the SPOT image, with had an overall accuracy of 80.6% and a kappa coefficient of 0.79. The subsequent analysis of the correlation between classification-derived housing sizes and the poverty showed 𝑅2 value of 0.33 and 0.25 respectively, with the average income and national poverty rate. In conclusion, a map showing the areas of poverty was produced by GIS analysis categorizing the areas on the basis of average family income, family size, and the percentage of small houses as poverty-related factors. This research will help officials, charities, decision-makers, and planners to focus their development efforts on the areas of poverty. Moreover, the results will be supporting the new Vision 2030 of Saudi Arabia, which aims to improve the quality of life through upgrades to housing, healthcare, and educational opportunities. The research could be the basis for other future research studies and subsequent experiments, and should encourage researchers to take advantage of satellite images and spatial analysis techniques for other applications.

Year manuscript completed

2021

Year degree awarded

2022

Author's Keywords

Poverty, GIS, Remote Sensing, Machine Learning, Classification, Support Vector Machine, Maximum Likelihood, Random Tree.

Degree Awarded

<-- Please Select One -->

Department

<-- Please Select One -->

College/School

<-- Please Select One -->

Dissertation Committee Chair

Robin Q. Zhang

Thesis Advisor

Robin Q. Zhang

Committee Member

Jane Benson

Committee Member

Haluk Cetin

Committee Member

Fahad Abdulaziz Almutlaq

Document Type

Thesis

Download

Share

COinS