BYU

Abstract by Jared Rydalch

Personal Infomation


Presenter's Name

Jared Rydalch

Degree Level

Undergraduate

Abstract Infomation


Department

Computer Science

Faculty Advisor

Ryan Farrell

Title

Circle Intensity Ratio Descriptors

Abstract

Regions of images located on the edges of lighter and darker regions can be found with edge detectors. In many cases these edge points can be further classified and augmented by analyzing a circular region around this edge point. Circle Intensity Ratio Descriptors are features generated on an image based on the ratio of “light” pixels to “dark” pixels in a circular area surrounding a point. These descriptors can vary in radius size and complexity. The simplest descriptor measures the ratio of light to dark pixels in the entire circle. More complex descriptors can provide an array of ratios from different regions of the circular area. These arrays of ratios can be used for various purposes, including their original purpose: describing tiger stripe endpoints.