Notes

In 2018, there have been 23 cases of deaths of Asiatic Lions. The major cause for the death is attributed to the Canine Distemper disease and Babesiosis Infection. Out of 23 deaths, 4 deaths have been caused by Canine Distemper Disease and 17 by Babesiosis Infection.

In October last year, officials discovered a body of Solo, a young tigress, in Bandhavgarh National Park in India. A case of suspected poisoning, her death was a testament to how conventional wildlife conservation methods have failed.

Hundreds of animals lose their lives to varied causes– floods, fire, and poisoning to poaching, every year. In India alone, 74 animal species are critically endangered, and the country ranks on the global ‘shame list’. Many of these animals are keystone species whose extinction can collapse entire ecosystems and food cycles.

Conventional and exhausting methods of surveillance for wildlife conservation have failed us.

Existing and conventional wildlife monitoring systems and methods of surveillance are insufficient and inefficient. Fortunately, the application of technology in wildlife and biodiversity preservation comes as a ray of hope. Applications of Artificial Intelligence and Machine Learning can change the dynamics of the field in favour of these threatened species.

Artificial intelligence is useful in understanding the repeated behaviours in animals– reproduction patterns, foraging routes, and hunting habits. Every animal exhibits the same basic behaviour.

An algorithm fit for wildlife surveillance needs to be trained on a large volume of metadata, including millions of images, sounds, and routes.




Algorithms used for animal detection are similar to the 2D Bounding Box. Animals in a single class need to be more precisely identified. Semantic segmentation coupled with image annotation is the best and correct technique to approach counting the animals. The method makes multiple animals detectable by using computer vision algorithms to localise the object.