"Somebody who lives and works in that area could be surveilled in almost everything they do," she said. "That is a society that most of us do not want to live in."
Analyzing video has been key to bombing investigations worldwide and was especially key to unraveling the Islamist bombings in July 2005 in London, perhaps the city with the world's most extensive network of video surveillance.
Cameras were key even as long ago as April 1999, when a bomb injured dozens of people at a market in Brixton, a district in south London, said Hugh Orde, a top police official during the investigation.
"We had, from a detective point of view, a disaster," he recalled.
But by reviewing surveillance videos, police identified a man wearing a cap low on his head and carrying a bag. In another clip, the bag was gone. The pictures broke open the case and led to an arrest.
Facial-recognition technology used by the FBI was first developed during the Iraq War to search for suspects in the aftermath of bombings, which typically featured crude, handmade explosives resembling those used in Boston.
"This is eerily similar to Iraq," said James Albers, senior vice president for government operations for MorphoTrust USA, based in Billerica, Mass., which makes facial recognition software for the FBI. Its development is part of a $1 billion push for a new generation of technology that eventually will include recognition of a person's irises and palm prints as well as traditional fingerprints.
Facebook and Google also have extensive facial-recognition data, gathered when users upload photos. Police typically need a search warrant for a particular suspect in order to access that information.
The facial-recognition software used by the FBI can lift shadows, sharpen blurry pictures and create three-dimensional images from flat ones, Albers said. It also analyzes each face to develop a unique "template" based on the shape, skin texture and distances among features.