Title :
Is color important for horizon line detection?
Author :
Gershikov, Evgeny
Author_Institution :
Dept. of Electr. Eng., Braude Acad. Coll. of Eng., Karmiel, Israel
Abstract :
Horizon line detection has several important applications in airborne and marine navigation as well as in surveillance and tracking. Many methods for horizon line detection use only the intensity component of still images and video sequences discarding the color information. Other algorithms locate the horizon line in color images using the information of all the color components. In this work we examine both groups of methods and we compare the performance of color based algorithms to that of luminance based methods. In particular, we compare several methods that have both color and grayscale versions. Our goal is to determine whether it is worthwhile to use color information in horizon line detection.
Keywords :
Hough transforms; aerospace computing; aircraft navigation; edge detection; image colour analysis; image sequences; marine engineering; marine navigation; object detection; object tracking; video signal processing; video surveillance; Hough transform; airborne navigation; color component information; edge detection; grayscale version; horizon line detection; horizon line location; intensity component; marine navigation; video sequences; Entropy; Gray-scale; Image color analysis; Image edge detection; Sea measurements; Transforms; Vectors; Hough transform; color importance; edge detection; global features; horizon line detection; local features;
Conference_Titel :
Advanced Technologies for Communications (ATC), 2014 International Conference on
Print_ISBN :
978-1-4799-6955-5
DOI :
10.1109/ATC.2014.7043395