Date of Completion
Dr. Reda Ammar, Dr. Chun-Hsi Huang
Field of Study
Computer Science and Engineering
Master of Science
Chemical detection is an important problem especially in the context of homeland security. We have designed an efficient algorithm to process complex multidimensional data to classify and detect harmful chemicals. The proposed solution is able to detect the chemicals with an accuracy of more than 90% which is better than the leading market solution.
Modern chemical detection technologies utilized in chemical defense suffer from poor accuracy. For example, they are limited in their broadband chemical detection performance (the ability to detect all threats) and are also unable to perform high confidence identification. The latter issue leads to false alarms as benign chemicals in the background environment are mistaken for a threat. Chemical detection is a time and data intensive problem. Development and implementation of effective detection algorithms is a big challenge.
In this thesis we study the problem of identifying chemicals and provide a novel solution. Our solution is based on characterizing the behavior of a chemical when a differential voltage (DF) is applied on Mass ion Spectrometry and studying the compensation voltage (CV) that is obtained. The CV voltage is represented as peaks where the chemicals exhibit high mobility. In the first phase, we present an innovative solution that identifies the chemicals uniquely with a high accuracy. In the second phase, we present an alternative approach to identify the peaks. The proposed algorithm can be used to obtain the peaks quickly without going through complex numerical procedures.
Periaswamy, Priya, "A Novel Clustering Technique to Identify Chemicals" (2014). Master's Theses. 660.
Dr. Sanguthevar Rajasekaran