DocumentCode :
3725118
Title :
Artifacts removal in LBP based flame edges using binary morphology
Author :
Zakir Hussain;Diwakar Agarwal
Author_Institution :
Department of E.C.E, G.L.A. University, Mathura, India
fYear :
2015
Firstpage :
184
Lastpage :
189
Abstract :
Edge detection plays an important role in flame and fire image processing. Analysis and monitoring of flame images has been done continuously to measure the visible area of a flame for providing the additional information such as fire region, effect, gases etc. An accurate value of an area could be obtained only if continuous and smooth boundary of flame shall be detected. Several traditional edge detection methods have been examined to determine flame edges but the results achieved does not focus on the continuity and clarity of the edges. Also, the Local Binary Pattern (LBP) based edge detection technique shows some artifacts (jaggies) in edges. In this paper, artifacts are removed by applying mmINTone (Mathematical Morphological INTerpolation in One step) magnification method which is based on binary morphology along with LBP based edge detection. The proposed algorithm detects jaggies in flame image and eliminates them, making the edges smoother. Experimental results of the flame image proved the effectiveness of the proposed algorithm.
Keywords :
"Fires","Image edge detection","Morphology","Transforms","Shape","Morphological operations","Interpolation"
Publisher :
ieee
Conference_Titel :
Signal Processing, Computing and Control (ISPCC), 2015 International Conference on
Type :
conf
DOI :
10.1109/ISPCC.2015.7375022
Filename :
7375022
Link To Document :
بازگشت