Научная визуализация, 2020, том 12, номер 5, страницы 61 - 78, DOI: 10.26583/sv.12.5.06
Optical Fiber Splicing Defect Segmentation Using Hybrid Active Contour
Авторы: A. Azizi1,A,B, Z. Azizi2,A
A Department of Electrical Engineering, University of Mohamed Khider, Biskra, Algeria
B Laboratory of Identification, Command, Control and Communication ‘LI3C’ University of Mohamed Khider, Biskra, Algeria
1 ORCID: 0000-0001-5151-8975, aziziabdallah87@yahoo.com
2 ORCID: 0000-0002-4632-1825, azizizineb93@gmail.com
Аннотация
In this paper, we are interested to segment splicing hot (infra-red) images using active contours in order to extract defects for the fiber splicing process on fiber splicer machine. The segmentation is assured by using a hybrid edge and region active contour; where region information uses local statistics to solve the problem of highlight or intensity inhomogeneity, whereas edge information based on Local Binary Patterns LBP, smoothes homogeneous regions and enhance contour information. Experimental results on hot images captured from the fiber splicer machine illustrate the effective performance of the proposed method.
Ключевые слова: Active Contours, Defect segmentation, Edge indicator, Fiber Splicing, Local binary pattern.
