The Utilization of SVM Algorithm and Deep Learning Methods to Identify Wild Animals

Identifying wild animals and being aware of their presence may serve a variety of purposes, the most significant of which is addressing the issue of accidents and collisions between vehicles and wild animals. There is a concern that this might endanger both people and animals. This study presents a problem involving the identification of objects by using three models of deep learning, namely VGG16, Inseptionv3, and MLP, as well as the support vector machine (SVM) technique. An analysis was performed on the data that was collected from 256 videos that included a variety of animals. Elk, Red Deer, Roe Deer, and Wild Boar are the four categories that are included in the data. We evaluated the outcomes of the four different algorithms and found that they were all equally effective. For VGG16, the highest accuracy was 99.41, and the validation scored was 98.85.