Object detection in a video based on frame differencing using deep learning
Authors
Varadharajan, Somasundaram
Issue Date
2019
Degree
MSc in Data Analytics
Publisher
Dublin Business School
Rights holder
Rights
Items in eSource are protected by copyright. Previously published items are made available in accordance with the copyright policy of the publisher/copyright holder.
Abstract
Earlier, security guards were employed to safeguard the people and premises. Due to advancements in technology and cheap storage cost, surveillance cameras have been installed in many public and private properties to prevent crime and theft activities. Though, it requires human intervention for monitoring and if any suspicious activities are detected it should be manually reported to the relevant authority. This is a time-consuming process and may have a high possibility of human error. Hence, there is a requirement to automate this entire process and a lot of research has been made to build a highly accurate model to overcome this major problem. This project is implemented to detect object like vehicles, human beings, animals and many other classes required for security monitoring purposes with the help of Convolutional Neural Network by using the frame differencing technique in the Structural similarity Image Metric algorithm. The built model has been trained effectively to detect around 80 objects present in the COCO dataset.