1. Detection and Classification of Animals using Machine Learning and Deep Learning
International Research Journal of Engineering and Technology (IRJET), January 2023
Authors: Shivam M. Butale, Kanchan D. Dongare, Suraj T. Sawant
Abstract: Human-Animal Conflict is a major problem where
an enormous number of resources is lost, and human life is in
danger. So, it is necessary to continuously monitor and prevent
animal intrusion. We aim to develop an animal detection and
identification system from images taken through monitoring
videos captured by motion-triggered cameras, called cameratraps. For these types of videos, existing approaches often
super from low detection rates due to low contrast between
the foreground animals and the cluttered background, as well
as high false-positive rates due to the dynamic background. To
address this issue, we will first develop an approach to
generate animal object region proposals using Advanced
Image Processing and then apply Artificial Intelligence
techniques to detect and classify the animal. We aim to
implement the animal detection and identification steps using
the following techniques: eXtreme Gradient Boosting
(XGBoost), Particle Swarm Optimization (PSO), Convolutional
Neural Network (CNN). We first determine if these region
proposals are true animals or background patches. We then
identify animals using the above algorithms. This Animal
Detection System can be used in a variety of applications like:
Animal Intrusion Detection in Industries set up in forest areas
/ remote location, in residential areas, Agriculture fields and
to avoid Animal-Vehicle collisions