{"id":{"repo_id":"rit","oai_identifier":"oai:repository.rit.edu:theses-8564"},"canonical_url":"https://search.ndltd.org/etd/rit/oai:repository.rit.edu:theses-8564","repository":{"repo_id":"rit","name":"Rochester Institute of Technology","base_url":"https://repository.rit.edu/do/oai/"},"display":{"title":"Image Processing Techniques to Separate Linear and Curvilinear Features in Textures","abstract":"Varied image processing techniques have been developed to extract or detect linear features from images. However, these techniques are targeted at extracting or detecting linear features, and it has been shown in an existing technique that the Fourier transform can be used in conjunction with the polar transformation to essentially lift or separate linear features from the background image. Extracting or detecting linear features in images involves locating these features in the image while separating or lifting them involves separating them from the background image such that we get two images: one image containing the linear features and the other image containing the background. This thesis presents approaches to separate linear and curvilinear features from textured backgrounds. The problem of separating linear features from a textured background is of importance in applications such as lithography, layout design and pattern recognition. The existing Fourier transform based approach of linear feature separation effectively separates randomly located lines that are spread throughout the entire image and is found to be ineffective when the linear features are of varied lengths and thickness. This thesis presents an approach to overcome this limitation of the Fourier transform based approach. This thesis presents two new window based techniques relying on the Fourier transform and the wavelet transform to lift randomly located lines of varied in lengths and thickness. The proposed techniques are built upon the existing Fourier transform approach. The performances of the proposed techniques are compared to the Fourier Transform approach through application to several images. It is observed that the proposed Fourier based block approach and wavelet based block approach consistently perform better than the existing approach. It is also observed that the proposed techniques effectively lift curvilinear features from textures too. The mathematical analysis and experimental results verifying this claim are presented.","abstract_html":"Varied image processing techniques have been developed to extract or detect linear features from images. However, these techniques are targeted at extracting or detecting linear features, and it has been shown in an existing technique that the Fourier transform can be used in conjunction with the polar transformation to essentially lift or separate linear features from the background image. Extracting or detecting linear features in images involves locating these features in the image while separating or lifting them involves separating them from the background image such that we get two images: one image containing the linear features and the other image containing the background. This thesis presents approaches to separate linear and curvilinear features from textured backgrounds. The problem of separating linear features from a textured background is of importance in applications such as lithography, layout design and pattern recognition. The existing Fourier transform based approach of linear feature separation effectively separates randomly located lines that are spread throughout the entire image and is found to be ineffective when the linear features are of varied lengths and thickness. This thesis presents an approach to overcome this limitation of the Fourier transform based approach. This thesis presents two new window based techniques relying on the Fourier transform and the wavelet transform to lift randomly located lines of varied in lengths and thickness. The proposed techniques are built upon the existing Fourier transform approach. The performances of the proposed techniques are compared to the Fourier Transform approach through application to several images. It is observed that the proposed Fourier based block approach and wavelet based block approach consistently perform better than the existing approach. It is also observed that the proposed techniques effectively lift curvilinear features from textures too. The mathematical analysis and experimental results verifying this claim are presented.","abstract_has_math":false,"creators":["Pidaparthi, Bhavani"],"institution":null,"degree_name":"Electrical Engineering (MS)","degree_level":"Thesis","degree_discipline":"Electrical Engineering (KGCOE)","degree_department":null,"school":null,"contributors":["Raghuveer Rao"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T09:01:00Z","subjects":[],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://repository.rit.edu/theses/7559","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://repository.rit.edu/theses/7559","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://repository.rit.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Arepository.rit.edu%3Atheses-8564","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Raghuveer Rao"]},{"key":"dc:creator","label":"Author","values":["Pidaparthi, Bhavani"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.created","label":"Dc Date Created","values":["2003-01-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical Engineering (KGCOE)"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Electrical Engineering (MS)"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://repository.rit.edu/theses/7559"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Varied image processing techniques have been developed to extract or detect linear features from images. However, these techniques are targeted at extracting or detecting linear features, and it has been shown in an existing technique that the Fourier transform can be used in conjunction with the polar transformation to essentially lift or separate linear features from the background image. Extracting or detecting linear features in images involves locating these features in the image while separating or lifting them involves separating them from the background image such that we get two images: one image containing the linear features and the other image containing the background. This thesis presents approaches to separate linear and curvilinear features from textured backgrounds. The problem of separating linear features from a textured background is of importance in applications such as lithography, layout design and pattern recognition. The existing Fourier transform based approach of linear feature separation effectively separates randomly located lines that are spread throughout the entire image and is found to be ineffective when the linear features are of varied lengths and thickness. This thesis presents an approach to overcome this limitation of the Fourier transform based approach. This thesis presents two new window based techniques relying on the Fourier transform and the wavelet transform to lift randomly located lines of varied in lengths and thickness. The proposed techniques are built upon the existing Fourier transform approach. The performances of the proposed techniques are compared to the Fourier Transform approach through application to several images. It is observed that the proposed Fourier based block approach and wavelet based block approach consistently perform better than the existing approach. It is also observed that the proposed techniques effectively lift curvilinear features from textures too. The mathematical analysis and experimental results verifying this claim are presented."]},{"key":"dc:description.note","label":"Dc Description Note","values":["<p>Physical copy available from RIT's Wallace Library at TA1637 .P53 2003</p>"]},{"key":"dc:title","label":"Title","values":["Image Processing Techniques to Separate Linear and Curvilinear Features in Textures"]}]}],"canonical_facts":{"dc:contributor":["Raghuveer Rao"],"dc:creator":["Pidaparthi, Bhavani"],"dc:date.created":["2003-01-01T08:00:00Z"],"dc:description.abstract":["Varied image processing techniques have been developed to extract or detect linear features from images. However, these techniques are targeted at extracting or detecting linear features, and it has been shown in an existing technique that the Fourier transform can be used in conjunction with the polar transformation to essentially lift or separate linear features from the background image. Extracting or detecting linear features in images involves locating these features in the image while separating or lifting them involves separating them from the background image such that we get two images: one image containing the linear features and the other image containing the background. This thesis presents approaches to separate linear and curvilinear features from textured backgrounds. The problem of separating linear features from a textured background is of importance in applications such as lithography, layout design and pattern recognition. The existing Fourier transform based approach of linear feature separation effectively separates randomly located lines that are spread throughout the entire image and is found to be ineffective when the linear features are of varied lengths and thickness. This thesis presents an approach to overcome this limitation of the Fourier transform based approach. This thesis presents two new window based techniques relying on the Fourier transform and the wavelet transform to lift randomly located lines of varied in lengths and thickness. The proposed techniques are built upon the existing Fourier transform approach. The performances of the proposed techniques are compared to the Fourier Transform approach through application to several images. It is observed that the proposed Fourier based block approach and wavelet based block approach consistently perform better than the existing approach. It is also observed that the proposed techniques effectively lift curvilinear features from textures too. The mathematical analysis and experimental results verifying this claim are presented."],"dc:description.note":["<p>Physical copy available from RIT's Wallace Library at TA1637 .P53 2003</p>"],"dc:identifier":["https://repository.rit.edu/theses/7559"],"dc:title":["Image Processing Techniques to Separate Linear and Curvilinear Features in Textures"],"thesis:degree_discipline":["Electrical Engineering (KGCOE)"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["Electrical Engineering (MS)"]},"updated_at":"2026-11-19T09:01:00Z"}