Date of Award

Summer 8-2016

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Mathematics

First Advisor

Harold E. Donley, Ph.D.

Second Advisor

Yu-Ju Kuo, Ph.D.

Third Advisor

Charles Lamb, Ph.D.

Abstract

A team of biological researchers wishes to investigate the eating behaviors of local insects to determine whether they prefer indigenous or invasive plant species. Current techniques involve researchers collecting leaf samples from the field and visually estimating the percentage of consumed area. This method is especially time consuming but may also be subject to a great deal of error due to individual bias and the erratic nature of human estimation. The task is to construct a program which accepts images of leaves, taken in a controlled environment, and returns estimates for the scale and angle at which the leaf occurs in the image. These values will then be used to limit the search space of the generalized Hough transform algorithm. The scale will be used to determine the missing leaf area. This would reduce the labor intensity of data collection and eliminate humans as a primary source of error.

Share

COinS