(CNN) -- The National Security Agency is reportedly capturing million of images per day to feed facial recognition programs.
Citing top-secret documents obtained by former NSA contractor Edward Snowden, the New York Times reported the agency 's reliance on such technology has grown in recent years.
New software allows the NSA to "exploit the flood of images included in emails, text messages, social media, videoconferences and other communications," the newspaper said.
"The agency intercepts 'millions of images per day' — including about 55,000 'facial recognition quality images' — which translate into 'tremendous untapped potential,'" the newspaper reported, citing documents from 2011.
Facial recognition is a computer-based system that automatically identifies a person based on a digital image or video source that is then matched to information stored in a database. The technology is powerful, but not always straightforward.
According to the New York Times: "It has difficulty matching low-resolution images, and photographs of people's faces taken from the side or angles can be impossible to match against mug shots or other head-on photographs."
The newspaper said it was unclear how many images have been acquired, nor was it clear how many people have been caught up in the program.
An NSA spokeswoman defended the program.
"We would not be doing our job if we didn't seek ways to continuously improve the precision of signals intelligence activities -- aiming to counteract the efforts of valid foreign intelligence targets to disguise themselves or conceal plans to harm the United States and its allies," NSA spokeswoman Vanee M. Vines said in response to the New York Times report.
"The lawful collection of foreign identity intelligence allows NSA to better identify and track such targets."
Speaking to the newspaper, Vines declined to comment on whether the agency collected facial imagery of American citizens from social media like Facebook.
CNN's Jim Stenman contributed to this report.