Over the past several years, nationwide attention has focused on using technology to deter terrorism; tighten U.S. borders; and allow local, state, and federal law enforcement agencies to share information. However, technology also has allowed police throughout the world to fight street-level crime. In many policing magazines, advertisements abound about technological advancements in internal affairs tracking software, communication hardware, automated fingerprint identification, voice stress analysis, and crime mapping.
One area of technology growing in leaps and bounds, however, has not garnered much attention. License plate recognition (LPR) software and hardware has gone from something only dreamed about in movies to a viable technological tool for local police and sheriff's departments. While certainly not perfect yet, the technology available today can be used to search for vehicles listed in AMBER Alerts, identify those driven by wanted persons, and recover ones reported stolen.
BACKGROUND
Known by many names, LPR technology employs cameras and computer software to discern the letters and numbers of vehicle license plates and then compares them with records contained in state and federal databases. While recognition software (used to "see" the letters of the license plate) has become increasingly more accurate over the past several years, early programs had severely low recognition rates. Initially, cameras used to obtain images of vehicles--and, therefore, license plates--had to be mounted in a fixed location. New technology allows the imaging cameras to be placed on the front or roof of a vehicle or in a patrol unit's light bar. Where LPR technology originally supplemented or replaced other identification devices (e.g., bar codes or radio equipment) to allow access and egress from secure facilities or charge fines or fees for travel or parking, new applications can enable the user to check information against department of motor vehicle or NCIC records.
Initially designed for use in parking lots (to record the time a vehicle entered), for access control (allowing authorized vehicles into a secure area), and for paying tolls, LPR technology recently has expanded into the realms of border control, identification of stolen vehicles, and traffic-fine enforcement (e.g., red-light running), with vendors marketing systems specifically for use by the law enforcement community. Combating auto theft represents a particularly applicable use of LPR technology. Law enforcement agencies throughout the United States constantly seek tools to locate and recover stolen vehicles, with the ultimate goal of making arrests that not only help solve open auto-theft investigations but other crimes as well. For example, crime analysts often track auto theft as a precursor to robberies.
LITERATURE REVIEW
As early as 2000, immigration officials in the United States and Canada began promoting LPR technology and cargo X-ray scanners. (1) In 2005, patrol stations along the U.S.-Mexico border also incorporated the technology to record the entry and exit of vehicles. (2) However, little academic research on LPR technology has occurred in the United States.
One of the most complete studies, conducted during a 4-month evaluation period in 2004, examined the effectiveness of the technology in identifying stolen vehicles and license plates, as well as vehicles driven by wanted felons. (3) While the review of the technology was generally favorable, the report indicated that the software produced over 1.8 million scans during the research period but caused 3,286 alarms, of which 108 were positive (meaning the license plate characters and state matched a valid entry in the computer system). Of particular interest, the study found that the reader could not match stacked or small characters on a license plate. At least one of the companies that produced the technology for LPR claimed that the device could prove instrumental in thwarting terrorist attacks. However, if the system cannot read stacked or small characters (often found on state commercial license plates), this issue becomes moot.
In Europe, the British Home Office also conducted research on the effectiveness of this technology. The pilot study showed that officers using the mobile plate recognition technology produced 100 arrests per officer per year--10 times the national average per officer. (4)
THE TECHNOLOGY
LPR technology involves cameras that feed information into a central processing unit (CPU) that then "reads" the license plate, converts it into optical character recognition (OCR), and then attempts to match it with "hot" plates listed in a state or national database. Currently, the systems usually do not conduct live inquiry into a government database. Instead, users download information daily to keep queries to NCIC and state databases at a minimum. The information is then sent to an agency-supplied mobile data terminal (MDT), usually a laptop. (5)
Types of Units
LPR units come in fixed configurations for mounting on light or sign posts for areas that pose special concerns for particular jurisdictions. Fixed-mount devices require a central dispatch site (possibly also a suitable place to store a central server for the LPR) to verify alarms from multiple locations. This proves crucial because false positive alarms may lead to sending officers on unnecessary calls. The personnel assigned to the fixed LPR conduct confirmation transactions, such as a criminal database check and a visual inspection of the plate photograph. The OCRs currently in use cannot discern license plates from different states or territories and also may read markings on commercial vehicles and confuse them with license plates. Additionally, some fixed-mount units rely on vehicle speeds of less than 35 miles per hour, so selection of an installation point becomes critical. Ideally, agencies should research suitable locations that can cover the traffic present. The more lanes of traffic monitored will require additional fixed-mount devices. LPR cameras typically can perform under low-light conditions, but, in some cases, agencies may encounter an additional cost of lighting the camera site to ensure the accuracy of data collection.
Mobile or portable LPR units also exist that can be mounted on marked or unmarked police, security, or government vehicles. Because of their portability, mobile devices allow for easy transfer from one vehicle to another. Some require installation onto a cruiser's light bar; others use magnets that temporarily attach to the roof. Each system has its applications, but end users should determine which unsecured loans format best suits their needs. The magnet-mounted device seems to work well with undercover tactical units as it can be easily transported and mounted on a variety of vehicles with minimal downtime. The portable system affixed to a marked vehicle's light bar allows consistent camera angles and familiarity with the equipment. In addition, covert devices that can operate in a variety of conditions and applications are now in production.
Data Input and Output
Getting information into the system is critical. Agencies may need the assistance of their criminal justice information administrators to receive daily lists of stolen vehicles, license plates, felony vehicles, and other relevant bad credit loans data. End users should determine a method of delivering that information to the LPR. Stand-alone units will require either a connection to the agency's network or wireless air card. The other option is a dedicated server connected to all devices, whereby information is collected in real time from respective criminal justice networks and then broadcast to the LPR. Because a combination of fixed and portable units could be deployed, the latter choice may prove the best for agencies with the financial resources to purchase both types. In addition, an important feature would be software that permits individual officers during their tour to input license plate information relevant to vehicles of interest. Because witness descriptions vary, the ability to enter partial license plate data would be useful, similar to entering "wild card" characters into the system.
No comments:
Post a Comment