سال انتشار: ۱۳۸۹
محل انتشار: ششمین کنفرانس ماشین بینایی و پردازش تصویر
تعداد صفحات: ۵
Adel Ghazikhani – PhD student, Computer Engineering Department, Ferdowsi University of Mashhad, Iran
Hamidreza Pourreza – Associate Professor, Computer Engineering Department, Ferdowsi University of Mashhad, Iran
We propose a novel algorithm for linear feature detection in images. Our algorithm is an extension to Beamlet Transform(BT). BT is a powerful transform for linear feature detection in images. It is based on the famous Radon transform. BT’s weakness is that it can’t detect lines with more than one pixel width. In this paper we propose an extension to BT that overcomes this problem in image processing applications. We compare the proposed algorithm with the original BT. The results show that the proposed algorithm has better results.