Date of Graduation

5-2017

Document Type

Thesis

Degree Name

Bachelor of Science in Industrial Engineering

Degree Level

Undergraduate

Department

Industrial Engineering

Advisor

Sullivan, Kelly

Reader

Rainwater, Chase

Second Reader

Runkle, Benjamin

Abstract

The agricultural industry’s continued success is critical to our future. However, many agricultural practices—specifically rice farming—are currently based on unsustainable methods. The processes used in rice farming consume water at an unsupportable rate, and, consequently, are serious contributors to rising emission levels. Several alternative growing methods have been introduced, and one of the most promising candidates is Alternative Wetting and Drying (AWD). An obstacle encountered by proponents of these methods comes in implementation. Studies have shown AWD is more successful in fields with higher leveling precision. The research in this study works to create a stepping stone for farmers to move to a more sustainable age of agriculture. This study proposes the creation of a dynamic database containing fields and their levelling technique and employs machine learning image classification methods to begin this work. If this database were to be created, effective programs could be developed to incentivize the adoption of alternative growth techniques. Through the use of convolutional neural networks, this research developed models to accurately classify satellite imagery of rice fields according to their levelling technique. It is our hope this research will be both expanded upon in the field of machine learning and used to effectively focus sustainable farming efforts.

Share

COinS