AI to Help Analyze Military Satellite Photographs

This post is also available in: עברית (Hebrew)

With the phasing out of the MiG-25 high altitude tri-sonic reconnaissance aircraft that could overfly enemy territory, India now relies heavily on its own expansive fleet of satellites for strategic reconnaissance or surveying areas of interest, such as in support of Indian Air Force strikes on terrorist targets in Pakistan.

India’s Armed Forces have recently initiated a project to adapt artificial intelligence for identifying and indexing military objects in aerial photographs or videos.

At present, identifying and locating military objects and potential targets in overhead surveillance pictures is done manually by intelligence personnel trained in imagery interpretation. While satellite imagery had played a crucial role in planning and executing Indian Air Force missions, post-strike satellite images of the area and other details, though claimed to be in their possession, were not released by the government, giving rise to speculation and conjectures over the effectiveness of the mission.

“Use of artificial intelligence and deep learning could be a powerful tool to enhance effectiveness and assist field commanders in decision making by providing imagery analysis in real time and suggesting suitable courses of action in a particular situation,” an officer said

Under the project, artificial intelligence and deep learning technology would be used to develop engines and algorithms to analyse satellite photographs as well as video feeds from unmanned aerial vehicles (UAV) for detection, classification, pattern recognition and decision support in real time, sources said.

Artificial intelligence and deep learning technology would be used to develop engines and algorithms to analyse satellite photographs as well as video feeds from unmanned aerial vehicles for detection, classification, pattern recognition and decision support in real time, according to tribuneindia.com.