Climate change has greatly impacted the quality of natural resources and also lead to the degradation of our planet’s ecological biodiversity. In response to such a critical issue, there are wildlife conservation programs worldwide. However, no effective methodology exists to assess the current biodiversity of an ecological area in real-time, which could help in environmental planning and track the changes in the place's biodiversity. Current ecological studies lead to the census of individual species by capture, mark, and recapture technique, which is cumbersome, inefficient, and time-consuming. Hence in this current research, we present the methodology in developing a real-time biodiversity map using deep learning algorithms like YOLOv3, which is applied on high spatial resolution photographs or real-time videos captured by inexpensive cameras mounted on a UAV platform for in-situ data acquisition. Such algorithms demonstrate a mean average precision of 95% within 6000 training iterations, which could help to transform wildlife classification, poaching detection, species identification, and geo-localization of species where the entire mapped region can be transfigured to an activity heat map updated in real-time and easily scalable to a larger region.