Department of Informatics and Telecommunications
Πληροφορική

2016-11-23

2016

Panou Dimitra

Βασίλειος Γάτος, Ερευνητής, τμήμα Πληροφορικής και Τηλεπικοινωνιών ΕΚΕΦΕ Δημόκριτος
Σέργιος Θεοδωρίδης, Καθηγητής, τμήμα Πληροφορικής και Τηλεπικοινωνιών ΕΚΠΑ

Αφαίρεση περιθωρίου και διόρθωση παραμόρφωσης σε έγγραφα από κάμερα

Greek

Border noise removal and perspective correction in camera-based document images

The goal of this bachelor thesis is the creation of an application that rectifies a document’s perspective distortion and removes its borders. The documents used are pictures taken by a camera. Specifically, the implementation of this application uses two methods. One method is based on Harris corner detection algorithm, while the other one uses Hough transformation. The first method finds document boundaries by specifying document corners, whereas the second one detects the maximum edges that make up the image’s contour. We compare the two methods and show that the method that uses Hough transformation appears to be more effective. Based on the experiments conducted using each method, we extracted the best method and the most effective sets of parameters and used them in the final version of the application

Technology - Computer science

perspective correction, border removal, edge detection, corner detection

Yes

8

Yes

24

110

https://pergamos.lib.uoa.gr/uoa/dl/object/1325010/file.pdf

https://pergamos.lib.uoa.gr/uoa/dl/object/1325010