Synopsis

Bangladesh faces a lot of natural calamities every year. The coastal morphology of Bangladesh influences the impact of natural hazards on the area. Bangladesh suffers from floods, cyclones, storm surge, river bank erosion, earthquake, drought, salinity intrusion, fire and tsunami. Cyclones and floods particularly caused massive damages. Cyclones occurred in 1970, 1991, 2007 and 2009 and killed 364,000, 136,000, 3,363 and 190 respectively. Due to unforeseen circumstances and improper management, the country suffers a lot from human lives and damaged goods and properties. This situation can be significantly improved by proper identification of damaged areas after a natural disaster. Image processing techniques combined with deep learning can play a vital role in detecting those damaged areas and eventually help the overall rescue procedure. 

Relevance of the Topic

Bangladesh is a land of natural calamities. Flood, cyclone, drought, famine destroy life and property every year. People live here fighting against the frequent natural calamities. In recent years our country has experienced a great number of natural calamities. Hence, this topic is immensely important and relevant to the current scenarios and it is high time to produce a workable solution to search and rescue procedure of those affected areas.

Future Research/Scope

Build a deep learning model to detect various objects in affected areas.

Build an app that can send a message to corresponding authority automatically after analyzing damages.

Build a prediction model based on historical data. 

Skills Learned

• Basics of machine learning.
• Usage of object detection models such as YOLOv4.
• Deep learning techniques such as CNN.
• Different CNN architectures such as AlexNet, VGGNet, ResNet, etc.
• Usage of high level neural network APIs such as keras.
• Usage of different libraries such as Numpy, Matplotlib, etc.
• Usage of different model performance metrics such as precision, recall, f1 score, confusion matrix, etc. 

Relevant courses to the topic

• Artificial Intelligence (CSE422)
• Neural Networks (CSE425)
• Machine Learning (CSE427)
• Image Processing (CSE428)

Reading List

• Munawar, H.S., Ullah, F., Qayyum, S. and Heravi, A., 2021. Application of deep learning on uav-based aerial images for flood detection. Smart Cities, 4(3), pp.1220-1242.
• Pi, Y., Nath, N.D. and Behzadan, A.H., 2020. Convolutional neural networks for object detection in aerial imagery for disaster response and recovery. Advanced Engineering Informatics, 43, p.101009.
• Cao, Q.D. and Choe, Y., 2020. Building damage annotation on post-hurricane satellite imagery based on convolutional neural networks. Natural Hazards, 103(3), pp.3357-3376.
• Lazin, R., Shen, X. and Anagnostou, E., 2021. Estimation of flood-damaged cropland area using a convolutional neural network. Environmental Research Letters, 16(5), p.054011.
• Pantaleoni, E., Engel, B.A. and Johannsen, C.J., 2007. Identifying agricultural flood damage using Landsat imagery. Precision Agriculture, 8, pp.27-36.