How facial recognition solves cases in Indiana
The state fusion center helps local agencies develop leads and catch criminals using facial recognition from Vigilant Solutions
Sponsored by Vigilant Solutions
By Tim Dees for PoliceOne BrandFocus
What if you could identify a suspect from a powerful network of databases instead of just a single database? The Indiana Intelligence Fusion Center uses facial recognition technology from Vigilant Solutions to assist local, state and federal agencies in generating leads to identify persons of interest and solve crimes, large and small.
Sgt. Jeff Carmin has spent most of his career in the more traditional roles of two city police departments and the Indiana State Police, performing traffic and criminal investigations throughout the state. He is now the director of operations at the Indiana Intelligence Fusion Center (IIFC) in Indianapolis.
What is a fusion center?
Fusion centers were established across the United States in the wake of 9/11 to create an information-sharing base among law enforcement agencies, whether federal, state, local or tribal. State and major urban area fusion centers provide information and intelligence to local agencies, working alongside federal agencies like the Department of Homeland Security and the Federal Bureau of Investigation to help local law enforcement.
The IIFC, operated by the Indiana State Police, leverages resources that would often be too costly for local agencies, said Carmin. It draws information from roughly 60 different databases and 72 other fusion centers around the United States.
The center also uses Vigilant Solutions for automated license plate and facial recognition technologies. Carmin said the center has used other facial recognition programs but none as successfully as FaceSearch from Vigilant Solutions.
“If you sent me a photo and the guy was looking to the side or down, or wasn’t looking straight on like a driver’s license photo, we weren’t getting good results,” he said. “What made us go with Vigilant, besides the cost, was the user-friendliness of it.”
Matching mug shots in the database
The IIFC’s experience with the Vigilant Solutions’ facial recognition system improved significantly after they uploaded mug shot photos stored in Indiana’s automated fingerprint recognition system into the comparison gallery maintained by Vigilant Solutions.
“Since then, we’ve had outstanding results when an investigator requests something via our facial recognition system,” Carmin said. “There are just a little under a million booking records.”
IIFC can also run comparisons against Vigilant Solutions’ master gallery, which contains records from various Vigilant users around the country. Other image databases are available for comparison as well.
“The Department of Homeland Security has a site with multi-state facial recognition. We also utilize that for 30 different states that can run it against their different databases, whatever they may be,” said Carmin.
How facial recognition technology works
Whereas humans recognize faces by an analog method, computers do it by digital mapping. By measuring and comparing distances and angles between facial landmarks – like pupils, the corners of the mouth, the tip of the nose, the lowermost aspect of the chin, etc. – a face can be expressed as a digital pattern the computer can compare with other patterns.
When a sample photo is submitted for analysis in a facial recognition system, the system locates records with a similar pattern in the database and returns them to the user as possible matches.
Local investigators wanting to make use of the IIFC’s facial recognition system upload photos or video still frames for comparison. An IIFC analyst enhances the photo, if necessary, using the Vigilant Solutions software, and the comparison results are sent back to the investigator, along with a log of enhancements made to the specimen image.
Vigilant Solutions’ FaceSearch system can correct for various problems, such as an off-axis angle or poor image quality.
“If the head is turned, you can actually get the head to turn to a more face-front type of photo,” Carmin said. “It will clear the photo up if it’s a blurry photo, it will add eyes if the eyes need to be added and that type of thing.”
Solving cases big and small
Carmin recounts one success story where the IIFC assisted in a case where $7,000 in gift cards was stolen.
“The only photos available were two profile shots, both taken from video surveillance footage,” he said. “Using the enhancement tools in the Vigilant software, an analyst was able to match the surveillance image to a booking photo of an inmate then housed in a county jail. The suspect was from Missouri and had no other ties to Indiana except that booking record. He also had outstanding arrest warrants.”
Thanks to the image match from the Vigilant system, the suspect was identified at the county jail, served with the warrants and charged with the gift card theft.
“We’ve had a lot of success with this system,” Carmin said. “It could be a homicide case, or it could range all the way down to a simple petty theft case.”
Using a powerful technology to solve a shoplifting case might seem like overkill, but it illustrates how making a tool like facial recognition available to all law enforcement can improve satisfaction for all concerned (the gift card thief being a likely exception).
About the Author
Tim Dees is a retired police officer and the former editor of two major law enforcement websites who writes and consults on technology applications in criminal justice. He can be reached at firstname.lastname@example.org.