z-logo
open-access-imgOpen Access
The application of UAV images in flood detection using image segmentation techniques
Author(s) -
Nur Ibrahim,
Siti Maryam SHARUN,
Muhammad Khusairi Osman,
Sherihan Mohamed,
Sharifah Hanis Yasmin Sayid Abdullah
Publication year - 2021
Publication title -
indonesian journal of electrical engineering and computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.241
H-Index - 17
eISSN - 2502-4760
pISSN - 2502-4752
DOI - 10.11591/ijeecs.v23.i2.pp1219-1226
Subject(s) - artificial intelligence , segmentation , cluster analysis , computer science , flood myth , computer vision , ground truth , rgb color model , image segmentation , remote sensing , pattern recognition (psychology) , geography , archaeology
The application of unmanned aerial vehicle (UAV) used to capture the images of the flood areas are becoming interest of most researchers recently. This is due to its versatilities of capturing the images with low-cost and real time responses. At present, the captured images are analysed manually by human experts, which cause the task labourous, time consuming and prone toerror. This study aims to develop an UAV-based automated flood detection system. Samples of images that consist of land and river areas were capture using a camera attached to UAV to emulate flooded and non-flooded areas.The RGB and HSI colour models were utilised to represent the flood images. Two image segmentation methods were studied, which are k-mean clustering and region growing. The segmented images were validated with manually segmented (ground truth) images. Simulation results show that the RG using gray images gave better segmentation accuracy (88%) as compared to the K-mean clustering (76%). Finally, an automated flood monitoring system based on the region growing method, called flood detection structure (FDS) was developed to detect and analyse the flood severity.

The content you want is available to Zendy users.

Already have an account? Click here to sign in.
Having issues? You can contact us here