Development of a wildlife sighting report and alert system using computer vision

Mohd Nadzri, Muhammad Syafiq (2021) Development of a wildlife sighting report and alert system using computer vision. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Development of a wildlife sighting report and alert system using computer vision.pdf - Submitted Version Download (443kB)
	Text (Full Text) Development of a wildlife sighting report and alert system using computer vision.pdf - Submitted Version Restricted to Repository staff only Download (3MB)

Abstract

The incident of wildlife entry on the public places is increasing nowadays. This is because animals require migration from place to place between specific causes. Therefore, this project allows us to overcome this problem. This project is called the Development of a wildlife sighting report and alert system using computer vision. This project will monitor wildlife activities if they are found around the scene. This project uses a system called Tensorflow which serves to store pictures of the physical characteristics of the wild animals and compare them with the current pictures captured. When recognition is detectable, the system will warn people near the scene by sending messages to their smartphone and another warning system will execute at the scene which is an alarm light with sound warning will be turned on. In this report, the list all of the software and hardware requirements have been clarified. This project able to recognize three types of wildlife which is wildboar, tiger and elephant with the accuracy of more than 50% on each condition. Raspberry pi 4 modules also have been explored in terms of using python, opencv and GPIO pin.

Actions (login required)

View Item