Open AccessModified Edge Detector for Coloured Images
Author(s) -
Ban Abdul-Rahman Al-Jubaury,
Luay Adware,
Laith Abdul-Aziz
Publication year - 2002
Publication title -
al-mağallaẗ al-ʿirāqiyyaẗ li-l-ḥāsibāt wa-al-maʿlūmāt/iraqi journal for computers and informatics
Language(s) - English
Resource type - Journals
eISSN - 2520-4912
pISSN - 2313-190X
DOI - 10.25195/ijci.v40i1.225
Subject(s) - sobel operator , rgb color model , hue , operator (biology) , artificial intelligence , histogram , computer vision , laplace transform , mathematics , edge detection , laplace operator , enhanced data rates for gsm evolution , computer science , pattern recognition (psychology) , image (mathematics) , image processing , mathematical analysis , biochemistry , chemistry , repressor , transcription factor , gene
In this paper, a study of the role nf clour information in detecting the edges of an image was conducted Therefore, different colour spaces with components corresponding to the attribute tumance, hue, and saturation le., VHS, and LUVI were implemented.Two edge delection techniques were applied to each of above colour spaces. These techniques are Sobel operator d Noalincar Laplace operator. A proposed nonlinear Laplace operator was implemented, and in encouraging results indicated its better cticiency than the traditional Tuonlinear Laplace operator. Diffcrcnt approaches were utilized to aclect El threshold value either manually or automatically. The automatic selection depends on the calculation of mean colour gredient magnitude or on the accumulated histogram. A suggestec and implemented mechanism based on using two threshold boundaries to detect edges in colour spaces other than RGB, the results indicated an improvement in the resultant eden image.