{"id":{"repo_id":"siu-theses","oai_identifier":"oai:opensiuc.lib.siu.edu:dissertations-2248"},"canonical_url":"https://search.dev.ndltd.org/etd/siu-theses/oai:opensiuc.lib.siu.edu:dissertations-2248","repository":{"repo_id":"siu-theses","name":"Southern Illinois University","base_url":"https://opensiuc.lib.siu.edu/do/oai/"},"display":{"title":"Object tracking and fetal signal monitoring","abstract":"This dissertation focuses on detecting and tracking small objects in complex backgrounds and on classifying ultra-sound fetal movement signals. The first part of this dissertation focuses on detecting and tracking laser spots which are small objects in a video frames. Many factors make this particular problem extremely challenging. Examples include, irregular shapes, intensity variations, color variations, speckle noise, obstacles, and background complexity. The new method developed combines the advantages of pyramid Lucas-Kanade (PLK) optical flow and optimal recursive Bayesian filters for detection and tracking. The detection and tracking results of the developed method are compared with the Kalman filter, extended Kalman filter, unscented Kalman filter, and the particle filter. It is shown that the new method yields better results. Monitoring fetal movements is crucial for the early detection of abnormalities in the fetus. The second part of this dissertation focuses on segmenting and classifying fetal movements which are aquired using six ultra-sound sensors strapped onto the stomachs of pregnant women. The goal is to classify the signals into two categories: general and startle. Segmentation is conducted using a signal-energy based algorithm and the signals are classified using features extracted from the segmented signals. Three supervised classifiers, the k nearest neighbor, nearest mean, and a support vector machine, are implemented and their performances are compared. The results show that the method developed is able to effectively classify the dichotomous signals.","abstract_html":"This dissertation focuses on detecting and tracking small objects in complex backgrounds and on classifying ultra-sound fetal movement signals. The first part of this dissertation focuses on detecting and tracking laser spots which are small objects in a video frames. Many factors make this particular problem extremely challenging. Examples include, irregular shapes, intensity variations, color variations, speckle noise, obstacles, and background complexity. The new method developed combines the advantages of pyramid Lucas-Kanade (PLK) optical flow and optimal recursive Bayesian filters for detection and tracking. The detection and tracking results of the developed method are compared with the Kalman filter, extended Kalman filter, unscented Kalman filter, and the particle filter. It is shown that the new method yields better results. Monitoring fetal movements is crucial for the early detection of abnormalities in the fetus. The second part of this dissertation focuses on segmenting and classifying fetal movements which are aquired using six ultra-sound sensors strapped onto the stomachs of pregnant women. The goal is to classify the signals into two categories: general and startle. Segmentation is conducted using a signal-energy based algorithm and the signals are classified using features extracted from the segmented signals. Three supervised classifiers, the k nearest neighbor, nearest mean, and a support vector machine, are implemented and their performances are compared. The results show that the method developed is able to effectively classify the dichotomous signals.","abstract_has_math":false,"creators":["Bai, Yang"],"institution":null,"degree_name":"Doctor of Philosophy","degree_level":"Campus Only Dissertation","degree_discipline":"Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":["Gupta, Lalit"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2016,"date_issued":"2016-08-01T07:00:00Z","date_published":"2016-08-01T07:00:00Z","updated_at":"2026-07-24T04:35:08Z","subjects":[],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://opensiuc.lib.siu.edu/dissertations/1244","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Gupta, Lalit"]},{"key":"dc:creator","label":"Author","values":["Bai, Yang"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical and Computer Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Campus Only Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://opensiuc.lib.siu.edu/dissertations/1244"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["This dissertation focuses on detecting and tracking small objects in complex backgrounds and on classifying ultra-sound fetal movement signals. The first part of this dissertation focuses on detecting and tracking laser spots which are small objects in a video frames. Many factors make this particular problem extremely challenging. Examples include, irregular shapes, intensity variations, color variations, speckle noise, obstacles, and background complexity. The new method developed combines the advantages of pyramid Lucas-Kanade (PLK) optical flow and optimal recursive Bayesian filters for detection and tracking. The detection and tracking results of the developed method are compared with the Kalman filter, extended Kalman filter, unscented Kalman filter, and the particle filter. It is shown that the new method yields better results. Monitoring fetal movements is crucial for the early detection of abnormalities in the fetus. The second part of this dissertation focuses on segmenting and classifying fetal movements which are aquired using six ultra-sound sensors strapped onto the stomachs of pregnant women. The goal is to classify the signals into two categories: general and startle. Segmentation is conducted using a signal-energy based algorithm and the signals are classified using features extracted from the segmented signals. Three supervised classifiers, the k nearest neighbor, nearest mean, and a support vector machine, are implemented and their performances are compared. The results show that the method developed is able to effectively classify the dichotomous signals."]},{"key":"dc:title","label":"Title","values":["Object tracking and fetal signal monitoring"]}]}],"canonical_facts":{"dc:contributor":["Gupta, Lalit"],"dc:creator":["Bai, Yang"],"dc:description.abstract":["This dissertation focuses on detecting and tracking small objects in complex backgrounds and on classifying ultra-sound fetal movement signals. The first part of this dissertation focuses on detecting and tracking laser spots which are small objects in a video frames. Many factors make this particular problem extremely challenging. Examples include, irregular shapes, intensity variations, color variations, speckle noise, obstacles, and background complexity. The new method developed combines the advantages of pyramid Lucas-Kanade (PLK) optical flow and optimal recursive Bayesian filters for detection and tracking. The detection and tracking results of the developed method are compared with the Kalman filter, extended Kalman filter, unscented Kalman filter, and the particle filter. It is shown that the new method yields better results. Monitoring fetal movements is crucial for the early detection of abnormalities in the fetus. The second part of this dissertation focuses on segmenting and classifying fetal movements which are aquired using six ultra-sound sensors strapped onto the stomachs of pregnant women. The goal is to classify the signals into two categories: general and startle. Segmentation is conducted using a signal-energy based algorithm and the signals are classified using features extracted from the segmented signals. Three supervised classifiers, the k nearest neighbor, nearest mean, and a support vector machine, are implemented and their performances are compared. The results show that the method developed is able to effectively classify the dichotomous signals."],"dc:identifier":["https://opensiuc.lib.siu.edu/dissertations/1244"],"dc:title":["Object tracking and fetal signal monitoring"],"thesis:degree_discipline":["Electrical and Computer Engineering"],"thesis:degree_level":["Campus Only Dissertation"],"thesis:degree_name":["Doctor of Philosophy"]},"updated_at":"2026-07-24T04:35:08Z"}