A Hybrid Model of Sectorization & Evacuation Path Detection for Disaster Affected Areas

Wagh, Vinaysheel Kishor (2020) A Hybrid Model of Sectorization & Evacuation Path Detection for Disaster Affected Areas. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Natural disasters’ effect such as damage to building’s hit by an earthquake can be detected and classified by monitoring satellite images. The application of image segmentation and classification to the analysis of natural disasters pose several challenges due to complex deep learning techniques and unavailability of data related to disaster prone area. This research proposes a novel method for sectorization and evacuation of natural disaster affected area by integrating image segmentation and classification method with a shortest path algorithm. The proposed method uses a public high-resolution satellite image dataset from Digital Globe. First the buildings are detected from the satellite images using a proposed segmentation model. Then the identified buildings are classified based on their incurred damage severity using a classification model. Finally, the best evacuation path is detected and recommended using the Dijkstra’s algorithm. The test set gives an F1-score of 0.84 for the segmentation model and 0.81 for classification model. The evacuation route detection model can generate dynamic evacuation route by analysing real-time satellite images of disaster affected area.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > QA Mathematics > Computer software T Technology > T Technology (General) > Information Technology > Computer software
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Dan English
Date Deposited:	21 Jan 2021 11:55
Last Modified:	21 Jan 2021 11:55
URI:	https://norma.ncirl.ie/id/eprint/4428

Actions (login required)

View Item