Video surveillance has traditionally been used a forensic tool, and less proactively at that. CCTV systems are generally used to record video and then the footage is scoured through to find the "what" that has happened. Of course, guards are sometimes used to view video walls live, but it becomes impractical to view more than a dozen or so live feeds a time. Video analytics have been developed to assist, or in some cases, replace guards in this task with Facial Recognition being one of the most fascinating of them all. However, most people are not aware of the history of Facial Recognition – a history that has been over 70 years in the making.
And it all began with Woodrow “Woody” Wilson Bledsoe.
After earning a PhD in mathematics from Berkeley, Woody Bledsoe began work at the Sandia Corporation in New Mexico where he first entered the world of computing and writing code, which then led to an interest in automated pattern recognition. Specifically, his interest resided in machine reading – “The process of teaching a computer to recognize unlabeled images of written characters.” Working with his colleague Iben Browning, the two invented what would become known as the “n-tuple” method, “a process that helped define the field of pattern recognition.” It worked like this:
They started by projecting a printed character—the letter Q, say—onto a rectangular grid of cells, resembling a sheet of graph paper. Then each cell was assigned a binary number according to whether it contained part of the character: Empty got a 0, populated got a 1. Then the cells were randomly grouped into ordered pairs, like sets of coordinates. (The groupings could, in theory, include any number of cells, hence the name n-tuple.) With a few further mathematical manipulations, the computer was able to assign the character’s grid a unique score. When the computer encountered a new character, it simply compared that character’s grid with others in its database until it found the closest match.
In 1960, Bledsoe and Browning founded their own company: Panoramic Research Incorporated. Their goal was to get a computer to recognize ten different faces, and they struggled from 1963 to 1965 to get a computer to recognize even one. Finally, they attempted to use what is known as the Bertillon method to help with pattern recognition. They used 122 photos representing 50 different people, took 22 measurements of the face on each photo, and wrote a program to process them. “At the end of the experiment, the computer was able to match every set of measurements with the correct photograph… [proving] that the Bertillon system was theoretically workable.”
Though the experiment in 1965 was a success, it was not infallible. Bledsoe still had issues with computers recognizing a pattern if there was any distortion of the facial measurements, such as a smile might cause. In 1967, he left Panoramic and began working at the University of Texas at Austin, where his final contributions to pattern recognition occurred.
He developed 2 programs for a computer to memorize one version of a face and then use it to identify another version.
“With the first, known as group matching, the computer would divide the face into features—left eyebrow, right ear, and so on—and compare the relative distances between them. The second approach relied on Bayesian decision theory; it used 22 measurements to make an educated guess about the whole. In the end, the two programs handled the task about equally well. More important, they blew their human competitors out of the water… This was the greatest success Woody ever had with his facial-recognition research.”
Though he eventually stopped his work in pattern recognition, it established a groundwork for others to build on. For instance, “in 1973 a Japanese computer scientist named Takeo Kanade made a major leap in facial-recognition technology. Using what was then a very rare commodity—a database of 850 digitized photographs, taken mostly during the 1970 World’s Fair in Suita, Japan—Kanade developed a program that could extract facial features such as the nose, mouth, and eyes without human input.”
Today, many of the challenges Bledsoe faced no longer exist. Not only have the advances in technology made computers “effectively self-teaching,” but the multitudes of imagery, such as selfies on social media, make for an endless supply of data. “Given a few rudimentary rules, they can parse reams and reams of data, figuring out how to pattern-match virtually anything, from a human face to a bag of chips.”
The history of Facial Recognition is an interesting one, but the fascination only grows with the technology itself. With manufacturer partners such as Avigilon and Axis Communications, LONG can provide you with Facial Recognition technology to help secure your building. Reach out today with questions about the products or benefits of Facial Recognition.
Richard oversees the sales and operations aspects of LONG’s security solutions team. He earned a bachelor’s degree from the University of Denver and has an extensive background in electronic security. Prior to that, he owned a computer networking company that designed, installed and maintained computer networks for businesses. Richard is always looking for new ways to provide customers with cutting edge security solutions that help streamline their businesses and improve safety and profitability. He also enjoys ice hockey, collecting guitars, writing and performing music and working on his vintage car collection.