Date of Award:

8-2026

Document Type:

Thesis

Degree Name:

Master of Science (MS)

Department:

Computer Science

Committee Chair(s)

Xiaojun Qi

Committee

Xiaojun Qi

Committee

Mahdi Masrullah Al-ammen

Committee

Shuhan Yuan

Abstract

Wheat rust is a serious plant disease that can reduce crop yield and quality. In practice, the disease is often noticed only after visible symptoms appear, when some damage may already be difficult to reverse. This thesis studies whether drone-based imaging can help detect wheat rust earlier and more reliably in field environments. 

Unlike an ordinary color photograph, a hyperspectral image records reflected light at many narrow wavelengths. These measurements can reveal useful information about plant condition, but they are also high dimensional, noisy, and difficult to analyze when only a limited number of labeled field samples are available. To address this challenge, the thesis first identifies a small set of wavelengths that are especially informative for wheat rust and plant stress. It then uses a two-part model: one part analyzes spatial appearance from a pseudo-color image, and the other analyzes spectral patterns from a compact set of selected wavelengths. The two sources of information are then combined to make the final decision. 

The method was tested on a public drone hyperspectral dataset containing field regions labeled as healthy, rust, or other. To improve label consistency before training, the dataset was refined using a vegetation-based filtering step. Under a fair comparison in which competing methods used the same fixed eight-band input, the proposed method achieved the strongest overall average performance. When each comparison model was also evaluated under its own original input setting, the proposed method remained competitive and ranked second overall while also being more efficient than the strongest competing model in model-only latency and computational cost. 

Overall, the results suggest that carefully selecting informative wavelengths and combining spatial and spectral information can provide a compact, interpretable, and practical approach for early wheat-rust detection. In the long term, this kind of method may support more timely crop monitoring and more informed disease management in precision agriculture.

Download

Share

COinS

Copyright for this work is retained by the student. If you have any questions regarding the inclusion of this work in the Digital Commons, please email us at .