Satellite Image Edge Detection for Population Distribution Pattern Identification using Levelset with Morphological Filtering Process

1st Annual Applied Science and Engineering Conference

工业技术;自然科学

Harsiti^1 ; Munandar, T.A.^2 ; Suhendar, A.^1 ; Abdullah, A.G.^3 ; Rohendi, D.^4

Information System Dept., Information Technology Faculty, Universitas Serang Raya, Banten, Indonesia^1

Informatics Engineering Dept., Information Technology Faculty, Universitas Serang Raya, Banten, Indonesia^2

Electrical and Electronic Engineering Dept., Faculty of Technology and Vocational Education, Universitas Pendidikan Indonesia, Bandung, Indonesia^3

Computer Scince Education Dept., Math. and Naturan Science Faculty, Universitas Pendidikan Indonesia, Bandung, Indonesia^4

关键词: Distribution patterns;    Morphological filtering;    Morphological images;    Population distribution patterns;    Satellite image analysis;    Satellite images;    Segmentation methods;    Statistical datas;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/180/1/012064/pdf
DOI  :  10.1088/1757-899X/180/1/012064

来源: IOP

【 摘 要 】

Population distribution pattern is directly related with economic gap of a region. Analysis of population distribution pattern is usually performed by studying statistical data on population. This study aimed to analyze population distribution pattern using image analysis concept, i.e. using satellite images. Levelset and morphological image filtering methods were used to analyze images to see distribution pattern. The research result showed that Levelset and morphological image filtering could remove a lot of noises in analysis result images and form object edge contours very clearly. The detected object contours were used as references to recognize population distribution pattern based on satellite image analysis. The pattern made based on the research result didn't show optimal result because Levelset performed image segmentation based on the contours of the analyzed objects. Other segmentation methods should be combined with it to produce clearer population distribution pattern.