The use of biometrics in law enforcement has a long history. In the early 1900s, police officers began using fingerprint evidence to help investigations. Since then, increasingly sophisticated technologies have expanded how biometric data enables law enforcement to help build an evidentiary case against suspects. As biometric technologies continue to advance, law enforcement agencies create and benefit from databases that combine facial recognition with traditional fingerprint and palmprint data to aid in investigations.
How is facial recognition used in law enforcement?
Facial recognition technology (FRT) compares photographs or video of a person with facial images against others from a database. These technologies enable law enforcement to close open cases more rapidly by providing the potential to help trace someone’s activities and locations.
Some examples of use cases include comparing a:
- Suspect’s picture to ATM videos when tracking down fraudsters
- Known or suspected terrorist’s photo to police camera video
- A kidnapping victim’s picture to closed-circuit security footage
While talking to people who may have seen a suspect or victim remains valuable, people’s memories are notoriously inconsistent. Facial recognition technology supplements these testimonies with facial recognition evidence.
What is the Difference Between Face Detection and Recognition?
Face detection identifies the existence of a human face in a frame, image, or video. For example, it would be able to identify whether a person was present in a picture of a forest.
Facial recognition compares a specific person’s photograph to an image containing a person to potentially identify that individual. For example, it could compare a person’s driver’s license photograph to the picture of a person in a forest, then tell you the likelihood that the person in the forest is the same as the person in the driver’s license image.
Since facial recognition is more precise, law enforcement agencies can use it more effectively when locating a specific individual during an investigation.
Facial Recognition Incorporates Additional Measurements
Face detection technologies use artificial intelligence to understand the general features that a human face has. For example, the software learns to search an image for eyes, noses, and mouths where they are expected to be on a face.
Facial recognition technology begins with face detection and then supplements it with measurements about a person’s face to create a unique identifier for an individual. Some examples of these measurements include the distance between:
- Left and right eyes
- Eyes and forehead
- Eyes and nose
- Forehead and mouth
The software then aggregates all these features and correlates them to create the facial version of a fingerprint. These facial measurements allow for the accurate identification of individuals.
Since facial recognition incorporates unique metrics, it expands the data law enforcement agencies can use to identify and connect evidence to a specific person of interest.
Facial Recognition’s Use Cases Align with Law Enforcement Needs
Face detection enables technologies to automate certain processes. Some examples of how security uses face detection include:
- Tracking the number of people in a store
- Counting the number of attendees at an event venue
While face detection can be used to count the number of people, it lacks the ability to respond to law enforcement’s need for precision.
Some examples of how law enforcement may use facial recognition technology include:
- Identifying a person of interest making a purchase in a store
- Locating a potential suspect trying to hide in a crowd at an event venue
Since it collects, aggregates, and analyzes more information than face detection technology, law enforcement can use face recognition more effectively and efficiently during an investigation.
With facial recognition technology, law enforcement agencies may be better able to locate a specific person in a crowd, even when that person is attempting to evade detection.
What Are the Benefits of Facial Recognition Technology in Law Enforcement?
As facial recognition technology becomes more advanced, law enforcement agencies increasingly adopt it. In 2021, the U.S. Government Accountability Office (GAO) surveyed 42 federal agencies employing law enforcement officers, finding:
- Three agencies owned a facial recognition system
- 12 used another entity’s systems, including the Bureau of Alcohol, Tobacco, Firearms and Explosives (ATF), Drug Enforcement Agency (DEA), and U.S. Immigration and Customs Enforcement (ICE)
- Five owned a system and used another entity’s system, including the Federal Bureau of Investigation (FBI), Secret Service, and the Transportation Security Administration (TSA)
Additionally, the GAO found that three agencies used facial recognition technologies on images of the U.S. Capitol attack on January 6th.
Faster Investigation Times
In a connected world, facial recognition technology enables law enforcement agencies to complete investigations faster. Whether using a criminal database or comparing across images taken in public places, the technology automates many manual tasks to help identify potential suspects or missing persons more efficiently.
Close Cold Cases
Some criminals evade law enforcement for years, often meaning that as suspects age, they become harder to identify. Since facial recognition technology uses immutable data points, it can help to identify people even if some features, like hair color, change. In 2017, the FBI used facial recognition technology to help identify a criminal after a 16-year search.
Rapid In-the-Field Identification
With the right facial recognition technology, law enforcement can reduce the time it takes to check a suspect’s identity against State and Federal databases. With this information, officers can determine whether someone they are talking to might:
- Have outstanding arrest warrants
- Have previously committed a crime
- Attempt to use a fake identity
What Law Enforcement Agencies Should Look for in a Facial Recognition Technology and the Use of a Biometric Identification System
While facial recognition technology is useful, not all law enforcement agencies can use it and its associated records in the same fashion. A biometric identification system supports fingerprint, face, and iris recognition to aid in biometric identification. A biometric search system enables matching a live sample against many existing biometric templates to find a record of a particular individual and verify their identity. The right biometric identification system helps to solve crimes.
When seeking an automated biometric identification system, law enforcement agencies should consider certain qualities to make the best choice for themselves.
Ease of Use for Records
Law enforcement officers collect investigation records in various ways, and their biometric identification system should easily accept those record types. An agency should look for a biometrics platform that incorporates:
- Livescan system interface: ensure the identification system interfaces with a variety of providers for the transmission of fingerprint, palm print, and mugshot photos recorded during booking
- EBTS Imports: consider if the identification system is FBI/NIST EBTS-compliant and can import EBTS records from all NIST-compliant systems
These features provide flexibility that gives law enforcement agencies a better return on investment.
Aggregated and Correlated Biometrics
The more data you feed an identification solution, the better it works. When considering a facial recognition technology solution, law enforcement agencies should look for ones that aggregate and correlate a complete selection of biometric information, including latent prints, palm prints, and facial images.
Since law enforcement officials can compare various biometric data types, they gain greater confidence in the evidence, enabling them to complete investigations faster.
Facial recognition technologies should be able to match facial features accurately, even in suboptimal situations. People try to evade detection by engaging in criminal activity in poorly lit areas or obfuscating their faces. To be effective, facial recognition technology must account for these situations. Additionally, it should incorporate how a face ages over time.
While witnesses may not recognize suspects as they age, technologies don’t have this problem, enabling law enforcement agencies to use them when investigating cold cases.
Speed of Search
Investigations are time-sensitive, so facial recognition technology needs to be fast. As an agency scales its use, a biometric identification system should be able to scale in response. A distributed search process across multiple processors or networked servers enables agencies to maintain a rapid search speed even as they add more records to their internal database.
Remote Search Capabilities
Only searching an owned database can create limitations. Law enforcement agencies should look for a biometric identification system that provides secure remote search functionality across local, state, and federal databases to gain all the efficiencies that facial recognition technology can provide. This enables agencies to gain insights into other crimes that someone might have committed, including lower-level offenses not reported to the state.
To learn more about how the right biometric identification and latent print analysis system can help your agency, read more about the benefits of AFIX Tracker here.