Ανάπτυξη κωδικοποίησης για αυτόματη αναγνώριση στόχων και χρήση στο λογισμικό VisualSFM

Μανωλόπουλος, Κωνσταντίνος; Manolopoulos, Konstantinos

dc.contributor.author	Μανωλόπουλος, Κωνσταντίνος	el
dc.contributor.author	Manolopoulos, Konstantinos	en
dc.date.accessioned	2018-03-22T11:11:12Z
dc.date.available	2018-03-22T11:11:12Z
dc.date.issued	2018-03-22
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/46758
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.15060
dc.rights	Αναφορά Δημιουργού-Μη Εμπορική Χρήση 3.0 Ελλάδα	*
dc.rights.uri	http://creativecommons.org/licenses/by-nc/3.0/gr/	*
dc.subject	Visualsfm	en
dc.subject	Sfm	en
dc.subject	Mvs	en
dc.subject	Coded	en
dc.subject	Target	en
dc.title	Ανάπτυξη κωδικοποίησης για αυτόματη αναγνώριση στόχων και χρήση στο λογισμικό VisualSFM	el
heal.type	bachelorThesis
heal.classification	Φωτογραμμετρία	el
heal.classificationURI	http://data.seab.gr/concepts/3d45aafccf6809ad356fb62c8abb5f1a03506430
heal.language	el
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2017-11-21
heal.abstract	Δημιουργία αλγορίθμου σε γλώσσα προγραμματισμού Python για αυτόματη αναγνώριση στόχων και χρήση στο λογισμικό VisualSFM	el
heal.abstract	The present diploma thesis aims at developing an appropriate encoding for automatic photogrammetric ta rget recognition and at its subsequent connection to the open source software VisualSFM, developed by Changchang Wu. This software gives the option for manual specification of common points of interest among images. The aim of this diploma thesis is to automate this process, which so far is not available in this software. Solving the correspondence problem between homologous points among images is of high importance in the applications of vis ion metrology. Along with the solution of this problem, accurate and error – free measurements of points of common interest ( GCP’s, tie points) among images, are of high importance . For this purpose, photogrammetric targets are used and placed in the scene , which usually represent points with know n ground coordinates. With the automatic recognition of these targets, many basic close – range photogrammetry and vision metrology application s such as georeferencing and scaling of the object are solved with high accuracy and speed. In particular, the coding systems for the recognition of these targets, developed and currently used, are divided into three main categories: i) dot distribution ii) concentric rings and iii) colour targets. Independently of the cate gory, every target should satisfy some specific preferences. Encoded markers are widely used in SFM – MVS applications. In the present diploma thesis, the strategy used is described in detail . This strategy was programmed in Python programming language in conjuction with the use of OpenCV library and was based on the photogrammetric targets which the Photoscan software uses. The methodol ogy was based on that used by Zhou and his colleagues for the self - calibration of a remote - sensing camera of Leica's MS50 total station. The final output file of the process contains the image coordinates and the ID’s of the targets and it is in a recognizable format to VisualSFM. Finally, to assess the performance and the accuracy of the developed algorithm, 6 photogram metric targets were placed on the surface of a frame, which offered a complex background to the tagets. Also, a series of experiments in different situations took place.	en
heal.advisorName	Γεωργόπουλος, Ανδρέας	el
heal.committeeMemberName	Ιωαννίδης, Χαράλαμπος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Αγρονόμων και Τοπογράφων Μηχανικών. Τομέας Τοπογραφίας. Εργαστήριο Φωτογραμμετρίας	el
heal.academicPublisherID	ntua
heal.numberOfPages	110 σ.
heal.fullTextAvailability	true