Image recovery and segmentation using competitive learning in a neighborhood system

dc.creatorLi, Chengcheng
dc.date.available2011-02-19T00:46:55Z
dc.date.issued2002-12
dc.degree.departmentComputer Scienceen_US
dc.description.abstractImage restoration and segmentation are important image processing techniques. In recent years, many researchers in the image restoration field have based their research methods on calculus of variation and mathematical statistics and do not directly incorporate the observed principles of a low-level animal vision. In a previous work, based on the principle of low-level mammalian visual system that deals with image restoration and segmentation problems from a more direct and easily understandable and acceptable aspect, a new algorithm incorporating competitive leaming method was developed. This algorithm yields improved performance over previous studies in synthetic image restoration. This paper furthers the development and application of this algorithm. This paper has purpose that is threefold. First, this paper presents results for reconstruction and estimation of uncorrupted images from a distorted or a noisy image by using competitive leaming method. This paper evaluates the CLRS (Competitive Leaming in image Restoration and Segmentation) method by experimenting with this algorithm on a variety of images and a wide range of parameters, both based on practices and theories. The meaning and value range of some parameters are discussed in detail. Second, we enlarged the size of the neighborhood used in CLRS to see the influence of neighborhood range. Third, we reviewed the current methods both in image restoration and edge detection, then we compared the restoration and segmentation results obtained from CLRS and all the other methods. The results showed that CLRS algorithm performances were consistently better or equal in edged preservation and comparable performance in enhancing within the boundaries. These results are based on simulation experiments on a set of synthetic and real images corrupted by Gaussian noise. We concluded that an interactive algorithm for image reconstruction and segmentation, CLRS, has been developed. This algorithm is based on the principle of competitive leaming.
dc.format.mimetypeapplication/pdf
dc.identifier.urihttp://hdl.handle.net/2346/22130en_US
dc.language.isoeng
dc.publisherTexas Tech Universityen_US
dc.rights.availabilityUnrestricted.
dc.subjectImage reconstructionen_US
dc.subjectImage analysisen_US
dc.subjectImaging systems -- Image qualityen_US
dc.subjectComputer algorithmsen_US
dc.titleImage recovery and segmentation using competitive learning in a neighborhood system
dc.typeThesis
thesis.degree.departmentComputer Science
thesis.degree.disciplineComputer Science
thesis.degree.grantorTexas Tech University
thesis.degree.levelMasters
thesis.degree.nameM.S.

Files

Original bundle

Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
31295017084392.pdf
Size:
13.34 MB
Format:
Adobe Portable Document Format