Show simple item record

dc.contributor.advisorXiong, Zixiang
dc.creatorMeng, Yan
dc.date.accessioned2016-04-06T16:19:13Z
dc.date.available2016-04-06T16:19:13Z
dc.date.created2015-12
dc.date.issued2015-12-02
dc.date.submittedDecember 2015
dc.identifier.urihttp://hdl.handle.net/1969.1/156220
dc.description.abstractIt is obvious that the excellence of education is boosted by the rise of E-learning, which provides electronic learning material through the Internet. The online education benefits more people than traditional classroom education by enabling anyone to participate in the learning process, regardless of the identity, location and personal schedule. The goal of lifetime study can be achieved in this way. In early years, the online lectures are recorded with static cameras placed at a fixed position, and result in visual effects of low quality. Currently, many online education systems employ operators to control static cameras or PTZ cameras to record lectures. The wages of operators increase the system cost, and the complex algorithm raises the computational power and slows down the running of the system. This thesis proposes a real-time face, upper body and writing detection and tracking system for online education. The recording equipment includes a PTZ camera and a static camera. It saves the hardware cost by lowering the number of cameras involved in the system and eliminating the need for operators. The controlling software makes use of OpenCv library to simplify and accelerate the image processing and execute face detection. The proposed system implements an efficient motion detection algorithm to detect writing action and rigorous controlling logic to switch views between the instructor and the board. The proposed system is thought to be economical from the perspective of hardware and efficient and accurate from the perspective of software. The recording is capable of keeping pace with the instructor and switching the view of the instructor to the view of the content written on the board smoothly as if an actual operator controls the camera. The proposed system brings competitive results regarding detection and tracking performance compared with conventional systems.en
dc.format.mimetypeapplication/pdf
dc.language.isoen
dc.subjectmotion detectionen
dc.subjectwriting trackingen
dc.subjectface detectionen
dc.subjectface trackingen
dc.titleA Real-Time Face and Writing Tracking System Based on the Combination of a PTZ Camera and a Static Cameraen
dc.typeThesisen
thesis.degree.departmentElectrical and Computer Engineeringen
thesis.degree.disciplineElectrical Engineeringen
thesis.degree.grantorTexas A & M Universityen
thesis.degree.nameMaster of Scienceen
thesis.degree.levelMastersen
dc.contributor.committeeMemberNeto, Ulisses Braga
dc.contributor.committeeMemberJiang, Anxiao
dc.contributor.committeeMemberSerpedin, Erchin
dc.type.materialtexten
dc.date.updated2016-04-06T16:19:13Z
local.etdauthor.orcid0000-0002-4173-9691


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record