All Eyes on the Border
Situation awareness critical to flexible response
- By Willem Ryan, Bob Banerjee
- Mar 03, 2009
Spread throughout the United States are 325 ports of entry that require surveillance and protection. Every year, Customs and Border Protection processes nearly a half billion people, along with 130 million vehicles and another 20 million cargo containers. Based on a five-year plan developed by CBP, the organization is working with federal, state, local, tribal and other entities to secure borders.
Among its five goals is to develop focused security for all aspects of port-of-entry operations. This includes enhancing situational awareness and creating flexible response capabilities.
Under the Secure Border Initiative, both state and national programs are working to deploy cameras, software and sensors to better track friendly and unfriendly entities. Security technology is advancing at a rapid pace, and many of the recent developments are ideal for helping CBP and other agencies with protection. These innovations revolve around capturing and recording high-quality video in challenging circumstances, including in remote border situations as well as at busy crossing areas, where hundreds of cars pass by every day.
Surveillance Between Ports of Entry
Between the hundreds of official POEs into the United States, there are thousands of miles that require a watchful eye as well, including 150 miles of a recently installed pedestrian fence and 115 miles of vehicle fencing on the southwest border.
With so many miles to cover and many of them situated in remote, rugged and inhospitable environments, cameras have become the eyes for border security. To cope with the geographic challenges of the more remote border locations, there are several ways to record and view video.
Night vision is critical in many border situations. Both passive and active night-vision options are available. Passive applications usually consist of standard low-light cameras that produce usable images through long exposure times and automatic gain control methods. These types of cameras provide an acceptable level of performance using ambient light, such as that from the moon and stars, to illuminate the scene.
Thermal infrared also can capture images by reading the heat differentials of people or wildlife. This technology relies on light intensification cameras that operate by amplifying the very low light levels given off by starlight or other light sources. However, while passive night vision will capture images of something moving, such as a person or animal, it cannot present the image in a high-resolution format for clear identification.
An Active Approach
An alternative option is active night vision, which has the ability to capture high-resolution images and provides clear views of faces or license plates. Cameras equipped with active night vision are often tested against DCRI standards, developed by John Johnson of the Army Night Vision laboratory. High-quality night-vision cameras will be able to accomplish each level of surveillance outlined by Johnson as detection, something in the field of view is present or moving; classification, the ability to distinguish between human or animal; recognition, friend or foe classification based on appearance or items being carried; and identification, the human can be identified as a specific individual. By conforming to Johnson’s criteria, the cameras deliver outstanding video at any of the four levels of surveillance.
By using ruggedized day/night cameras with active infrared illumination, even the darkest spots along the border can be viewed and recorded with excellent picture quality. But, even with cameras that are capturing high-quality video covering hundreds of miles along the border, there are too many video streams for border patrol personnel to monitor effectively.
Intelligence can be added to these video systems through the use of IP cameras with embedded video content analysis capabilities or through rugged encoders that offer the same embedded feature. With video content analysis, CBP staff can better detect and act on suspicious behaviors. And by relying on analytics that runs at the edge, rather than on racks of centralized PC servers, end users are working on the raw video, drastically lowering the systems’ total cost of ownership.
VCA can be configured to detect a number of behaviors that would indicate a person is attempting to cross a border location. For example, the technology can detect when an object has been removed, such as a portion of a perimeter fence. Loitering is another behavior that cameras at borders can detect. VCA can be configured to detect if a person enters an area and doesn’t leave after a specified time, indicating a possible threat. In this case, the technology serves as an earlywarning system to alert operators that someone may be looking for an opportunity to cross a border point. With advanced warning, the CBP personnel can send a patrol to the area before the border is actually crossed.
Finally, one of the most common uses of VCA at a border is the tripwire function. Here, the camera is programmed to set an invisible line, likely at a border point that isn’t controlled by staff, such as a fence or along a waterway. When the camera detects that a person or object, such as a boat, crosses that line, it will issue an alarm to the operator and bring up video of the scene on the operator’s screen.
For the most part, these behavior-based sensors are able to distinguish among small animals, people and vehicles, thus reducing false alarms and providing greater perimeter protection.
While VCA can act as an assistive technology in improving security at border locations, the technology requires a high-quality picture to function properly at night. Illumination during dark hours is required for effective 24/7 surveillance, particularly in remote areas. Combining ruggedized cameras, active infrared illumination and video content analysis technology, CBP officials are better equipped to watch over and secure convenient crossing points in remote sites.
Recording Along the Border
Recording video at remote border locations also is important for evidentiary purposes and for helping to identify trends or events surrounding an incident through forensic video searches. And, while proper lighting is essential for effective surveillance, it is important for optimizing recording technology as well. Snowy images or video resulting from inadequate lighting produces larger file sizes because noise in the signal requires greater storage resources.
By improving video quality through the use of day/night cameras with infrared illuminators, video can be compressed further, resulting in smaller file sizes and reduced bandwidth requirements for transmission across a network. Active IR actually tricks the system into thinking it is daytime and thus avoids nighttime noise issues.
However, even with high-quality images that require lower bandwidth for streaming across the network, bandwidth in geographically remote areas is a precious resource. Locations are usually networked via a WAN, and the bandwidth capabilities are small, often because wireless connectivity is the only practical way to reach the site. Therefore, it isn’t practical to stream all video at a high resolution to a central location.
Instead, dual-streaming—where highquality video streams are produced for recording at the network’s edge and lowquality streams are produced for viewing over the WAN—can be used. High-resolution video can be recorded at each site through the use of a rugged encoder with embedded storage options, such as an automotive-grade internal 80 GB hard drive, an external USB hard drive or even a compact flash card that can store up to 16 GB of video recording.
Rugged encoders are a reliable option for recording in the inhospitable environments often found in remote border locations. They can typically operate in more severe conditions—from -22 to 140 degrees Fahrenheit. Compared to a PCbased NVR, which requires 60 to 70 degrees Fahrenheit for proper operation, the choice is clear for border points located in more extreme environments.
While points along the border may require the greatest security enhancements at this time, there are improvements that also can be made to fortify our official ports of entry. One additional means of improving awareness at official ports of entry is via the use of surveillance cameras designed for license plate capture. License plate images can be used to check for stolen vehicles or those tied to criminal activities, or to match them against a database to speed processing at the border.
Although a relatively new venture within the United States, license plate capture has been deployed in several countries, including China and Finland, to improve processing and to shorten wait time at borders. For example, Finnish customs installed Bosch REG license plate capture cameras to help with border traffic operations in the Karesuvanto region, which separates Finland from Sweden. The cameras send license plate images to license plate recognition software designed to process constant traffic over multiple lanes.
This example shows how effective license plate capture and recognition can be in even the most extreme conditions. In the Scandanavian region, the cameras are exposed to sub-zero temperatures, low- or no-light conditions during the winter months and extreme glare from the low position of the sun in the winter sky. With active IR illumination, the license plate images can be captured even in complete darkness, and ambient rejection technology overcomes reflections from wet roads and glare from headlights. The high-quality of the images results in outstanding recognition rates.
Through the use of license plate capture technology, Finnish customs has been able to effectively identify vehicles, helping to upgrade its security and border operations. Cameras used for license plate recognition should be dedicated to that task alone, rather than performing double duty for both image capture and general surveillance. And, it is ideal to use one camera per traffic lane.
Positioning the cameras also is critical to getting the best image capture. Ideally, a license plate should be viewed at no more than a 35-degree angle, and traffic should be managed so the camera has a clear view of the license plate area at all times. While the technology requires careful installation, there is no doubt about its ability to enhance security at official ports of entry along the border.
Overall, whether it comes in the form of license plate capture cameras or those with integrated active infrared illumination and video content analysis, modern surveillance technology presents many options for improving security at our nation’s borders. And, while there are many challenges unique to this type of application, there are nearly as many solutions to overcome them.
This article originally appeared in the March 2009 issue of Security Today.