Date of Award
Spring 1985
Document Type
Thesis - Restricted
Degree Name
Master of Science (MS)
Department
Biomedical Engineering
First Advisor
Doerr, Thomas A.
Abstract
Edge enhancement and detection algorithms are used as a preprocessing step for object recognition in digitized images. Edge enhancement techniques using Sobel and linear convolution operators are analyzed using several Computerized Tomographic (CT) images. A linear gray scale control image, a circular phantom, and a chest image are used for the enhancement analysis. Conditional averaging is used to reduce the effects of noise in the scanned images. An edge thinning and thresholding algorithm is performed on the images as a preprocessing step to edge detection. The edge thinning and thresholding technique is used to reduce multiple indications of a single edge within a specified range of edge boundaries. Edge detection, using a modified edge relaxation process, is performed on the edge thinned images. Edge detection provides a means of forming continuous edge boundaries using the context of the surrounding image. Images are processed using the edge detection techniques and the results are discussed.
Recommended Citation
Hawthorne, Jon S., "Edge Enhancement and Detection for Object Recognition in Computerized Tomographic Images" (1985). Master's Theses (1922-2009) Access restricted to Marquette Campus. 4787.
https://epublications.marquette.edu/theses/4787