University of Twente Student Theses


Crop Type Discrimination Using Field and Satellite Hyperspectral Measurements in Busia, Kenya

Mlawa, Kelvin Aslen (2023) Crop Type Discrimination Using Field and Satellite Hyperspectral Measurements in Busia, Kenya.

[img] PDF
Abstract:This study aimed to assess the potential of hyperspectral remote sensing data for distinguishing between maize and cassava crops in Busia County, Kenya, at both leaf and satellite levels. Through fieldwork and spectral analysis, the research sought to provide insights into hyperspectral-based crop discrimination. Field measurements, including crop spectral data, were collected in July and August 2022. Satellite data from PRISMA acquired in July 2022 was also used to analyze crop spectral reflectance. Statistical tests, continuum removal, and band depth analysis were applied to identify differences between cassava and maize crops at various wavelengths. The findings indicated significant differences in spectral reflectance between cassava and maize at both leaf and satellite levels, primarily in specific wavelength regions. Notably, the study successfully discriminated between the two crops at the leaf level with 94% accuracy and a kappa score of 0.89, and at the satellite level with 77% accuracy and a kappa score of 0.54. These distinctions were attributed to biophysical and biochemical variations between the two crops. The study recommends further testing of this method with other crop types to enhance hyperspectral data's utility in crop discrimination. This research underscores the potential of hyperspectral data for crop monitoring and management, aiding resource allocation and yield prediction in agriculture.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Subject:48 agricultural science
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page