Odisha Wildlife Sanctuary Deploys AI Cameras to Track Eight Animal Species

Debrigarh wildlife sanctuary in Odisha has become the first in the state to implement artificial intelligence-powered camera traps for wildlife monitoring, representing a major upgrade in how conservation officials track animal populations and movements.

The camera system uses a Convolutional Neural Network engine to automatically recognize eight species: gaur, leopard, sambar, spotted deer, wild pig, sloth bear, and nilgai. Unlike conventional anti-poaching cameras that focus on detecting human activity, this technology processes images on-device to identify wildlife, filtering out irrelevant data and eliminating the need for staff to manually review thousands of photos.

According to Anshu Pragyan Das, divisional forest officer of the Hirakud wildlife division, the technology complements traditional foot patrols by improving surveillance efficiency and enabling faster responses to potential threats. A central control room now monitors live feeds from the equipment in real time, allowing officials to track wildlife activity more effectively across the sanctuary.

The surveillance infrastructure includes an AI tower in the Chourasimal area with a data-capture radius of approximately 5-6 kilometers, covering roughly 70 square kilometers. Additionally, 18 AI camera traps have been positioned throughout a 2-3 kilometer forest section, spaced about 200 meters apart to monitor interior and remote areas.

The cameras operate through a wireless mesh network that allows devices to relay information to each other without cellular connectivity, addressing the challenge of limited mobile coverage in remote forest zones. Field personnel can remotely activate individual cameras on demand, while thermal imaging enables continuous monitoring during nighttime hours when standard photography is ineffective.