Less than two hours later, Henry Yamada plopped down with a thud in front of the round table, setting his laptop on it and immediately flipping up the screen. The table was large and old, covered in a light brown laminate dotted with a smattering of tiny nicks and scratches. Just like everywhere else in the world, the downtown Phoenix library was a shell of its former self. Still wall-to-wall with bookcases and books, but almost entirely devoid of actual people.
“Are you sure this is going to work?”
“No.” He began typing without looking back at Rachel. “But I’ve seen programs that do this.”
Rachel looked over as Waterman and Reiff both approached and sat down. Together, they dropped several old geographical magazines on the table and began leafing through them, stopping to study maps of the surrounding area.
Turning back to Yamada, she said, “So what do these programs do?”
“Ever hear of image stitching?”
“No.”
Henry paused and looked at the others. “It became big about forty years ago, based on something called ‘projective geometry.’”
“Kid,” said Waterman, “if you’re trying to get us to understand something, you’re going in the wrong direction.”
Yamada nodded. “The general concept is not as complicated as it sounds. It has to do with computer pixels, tiny dots that make up any given image or picture. Ranging from hundreds of thousands to several million. Millions of dots that together create an image and dots that can also act like tiny digital fingerprints. Especially when you group them! Because if you group enough, their colors and hues can create a combination of unique digital values. Unique enough that you can then use a mathematical algorithm to ‘search’ for that particular combination.”
“I didn’t understand any of that,” said Rachel.
Yamada grinned. “Let me try again. The pixels are not just unique in their colors and hues but also in what they are displaying. Say, for example, a building or piece of architecture. All the details of its shape; its edges, its lines, various shading, and angles. When you include everything, even the most boring object becomes unique. And you may not need the whole object to fingerprint it; a lot of times, you just need a piece of it. Nor does it even have to be an actual picture; it just needs enough detail for the computer to create that fingerprint when we scan it.” He returned to his keyboard and resumed typing. “A lot of data was lost from the internet, but there’s still a lot of stuff out there.”
Rachel was not following. “So, what then, this all comes down to what we can find?”
Yamada stopped again and shook his head. He then grinned and looked at Reiff. “No, it comes down to how well John can draw.”