Wildlife Monitoring System Using Machine Learning

Abstract

Camera trapping is the approach used to proceed and monitor wild animals by the collection of data at a low cost. Using machine learning algorithms to build an automatic image classification and identification algorithm. Monitoring wildlife is necessary for conservation. Camera trapping is the most commonly used technique for the monitoring of wild animals, in which camera trapping automatically detects the presence of the wild animals in the frame and which a huge volume of data is being stored. T two types of algorithms compare and find the high accuracy of the algorithm and give the best output result for camera traps for wild animal detection. Our current work aims in investigating the different algorithms in machine learning that includes the Random Forest algorithm (RFA), and the Support vector machine algorithm (SVM).In our study, the overall comparison is about the accuracy of the Random forest algorithm and the support vector machine algorithm models are being observed for the discussion. The outcome of the experiment suggests that the Random forest algorithm is more accurate than the support vector machine algorithm. The experiment uses a Yolo dataset that contains the various categories of animals among which 7 animals were selected to test the performance of the models.