Date of Award
5-2014
Document Type
Thesis
Degree Name
Master of Science (MS)
Legacy Department
Computer Engineering
Committee Chair/Advisor
Walker, Ian
Committee Member
Birchfield , Stanley
Abstract
Ear detection is an actively growing area of research because of its applications in human head tracking and biometric recognition. In head tracking, it is used to augment face detectors and to perform pose estimation. In biometric systems, it is used both as an independent modality and in multi-modal biometric recognition. The ear shape is the preferred feature used to perform detection because of its unique structure in both 2D color images and 3D range images. Ear shape models have also been used in literature to perform ear detection, but at a cost of a loss in information about the exact ear structure. In this thesis, we seek to address these issues in existing methods by a combination of techniques including Viola Jones Haar Cascades, Active Shape Models (ASM) and Dijkstra's shortest path algorithm to devise a shape model of the ear using geometric parameters and mark an accurate contour around the ear using only 2D color images. The Viola Jones Haar Cascades classifier is used to mark a rectangular region around the ear in a left side profile image. Then a set of key landmark points around the ear including the ear outer helix, the ear anti-helix and the ear center is extracted using the ASM. This set of landmarks is then fed into Dijkstra's shortest path algorithm which traces out the strongest edge between adjacent landmarks, to extract the entire ear outer contour, while maintaining a high computational efficiency.
Recommended Citation
Ravindran, Satish, "Ear Contour Detection and Modeling Using Statistical Shape Models" (2014). All Theses. 1992.
https://tigerprints.clemson.edu/all_theses/1992