With the advent of nanotechnology, nanomaterials have drastically improved our lives in a very short span of time. The more we can tap into this resource, the more we can change our
lives for better. All the applications of nanomaterials depend on how well we can synthesize the nanoparticles in accordance with our desired shape and size, as they determine the
properties and thereby the functionality of the nanomaterials. Therefore in this report, it is focused on how to extract the shape of the nanoparticles from electron microscope images
using image segmentation more accurately and more efficiently. By developing automated image segmentation procedure, we can systematically determine the contours of an assortment of
nanoparticles from electron microscope images; reducing data examination and interpretation time substantially. As a result, the defects in the nanomaterials can be reduced drastically by
providing an automated update to the parameters controlling the production of nanomaterials. The report proposes new image segmentation techniques that specifically work very effectively
in extracting nanoparticles from electron microscope images. These techniques are manifested by imparting new features to Sliding Band Filter (SBF) method called Gradient Band Filter (GBF)
and by amalgamating GBF with Active Contour Without Edges method, followed by fine tuning of μ (a positive parameter in Mumford-Shah functional). The incremental improvement in the
performance (in terms of computation time, accuracy and false positives) of extracting nanoparticles is therefore portrayed by comparing image segmentation by SBF versus GBF, followed by
comparing Active Contour Without Edges versus Active Contour Without Edges with the fusion of Gradient Band Filter (ACGBF). In addition we compare the performance of a new technique called
Variance Method to fine tune the value of μ with fine tuning of μ based on ground truth, followed by gauging the improvement in the performance of image segmentation by ACGBF with fine
tuned value of μ over ACGBF with an arbitrary value of μ. / A Thesis submitted to the Department of Industrial & Manufacturing Engineering in partial fulfillment of the requirements for the degree of Master of
Science. / Fall Semester 2015. / November 09, 2015. / Active Contours, Image Segmentation, Nanoparticles, Sliding Band Filter / Includes bibliographical references. / Chiwoo Park, Professor Directing Thesis; Abhishek Shrivastava, Committee Member; Tao Liu, Committee Member; Adrian Barbu, Committee Member.
| Identifer | oai:union.ndltd.org:fsu.edu/oai:fsu.digital.flvc.org:fsu_291349 |
| Contributors | Allada, Kartheek (authoraut), Park, Chiwoo (professor directing thesis), Shrivastava, Abhishek Kumar (committee member), Liu, Tao, 1969- (committee member), Barbu, Adrian G. (Adrian Gheorghe), 1971- (committee member), Florida State University (degree granting institution), College of Engineering (degree granting college), Department of Industrial and Manufacturing Engineering (degree granting department) |
| Publisher | Florida State University |
| Source Sets | Florida State University |
| Language | English, English |
| Detected Language | English |
| Type | Text, text |
| Format | 1 online resource (82 pages), computer, application/pdf |
Page generated in 0.0139 seconds