Wild Animal Detection using a Machine Learning Approach and Alerting using LoRa Communication
Gnaneshwar Bandari, Prof. L Nirmala Devi, P. Srividya
Abstract
It is crucial to effectively and consistently monitor wild animals in the vicinity of the forest boundaries. In this paper, an algorithm is demonstrated to identify wild animals in order to protect them. Since there are so many different kinds of animals, manually recognizing them might be challenging. To better effectively monitor animals, this system categorizes them based on their images. Animal monitoring, theft prevention, and animal-vehicle accident prevention can all be aided by animal detection and classification. Applying efficient deep learning techniques can help with this. When an animal is detected, an alert will be sent to another device through LoRa communication because standard communications like WiFi and GSM technologies may not be available in the remote sensing area. Computer vision using Python is utilized for image processing. Raspberry Pi development board is used where the model is deployed and live streaming is processed to detect a wild animal which will differentiate between Humans and Domestic animals. Once, a wild animal is detected an alert is sent through LoRa communication to an end-user who is monitoring the wildlife.