Project Overview

Convergence of Jump Based Binarization Algorithms

Faculty Sponsor

Hiva Samadian (hsamadian@colgate.edu)

Department(s)

Computer Science
Mathematics

Abstract

A Binarization algorithm is a process for digitalizing a given data. It has applications in subjects such as image processing and gene expression. Some binarization algorithms binarize the input data by finding a numeric value called threshold. The variation on the input data causes the variation on the returning threshold. It is important to characterize the behavior of binarization algorithms and see how stable is the algorithm's threshold. The goal of this project is to see if the sequence of thresholds produced by a given algorithm when applying on a sequence of data converges or not. 
The objective for this summer is to investigate a jump based algorithm using an already established analytical framework. This may involve adjusting the analytical approach for the algorithm under study. The analytical results have to be tested by experiments on the real gene expression data. During this project, you will (1) get acquainted with the literature and subject of the research (2) if new to Matlab, learn to read and write code in Matlab (3) apply your knowledge that you have learned in courses such as logic or discrete structure on a real-world problem (4) read, understand, and write critique on a research paper and learn about some other general components of doing research.

Student Qualifications

Being familiar with mathematical logic and programming with Matlab is desired.

Number of Student Researchers

1 student

Project Length

9 weeks


Applications open on 01/03/2020 and close on 02/05/2020


<< Back to List





If you have questions, please contact Karyn Belanger (kgbelanger@colgate.edu).