Novel vision system autonomously detects and identifies drone threats – Military & Aerospace Electronics | Ad On Picture

ATLANTA – Commercial and recreational drone use has skyrocketed in recent years as the technology becomes cheaper and easier to use. From infrastructure inspection and emergency response to filmmaking and amateur photography, drones have a wide and growing range of uses. Unfortunately, drones can also pose threats, whether intentional or not. In places where drones pose a threat, such as airports, prisons, border patrols, and military scenarios, airspace security becomes paramount.

Luckily, as drone technology has evolved, so have drone detection capabilities. Typically, high-frequency sensors, radar, acoustic sensors, cameras and other technologies are used for such tasks. Some technologies even offer autonomous capabilities.

Drone threats are emerging

In the United States alone, the number of drones registered with the Federal Aviation Administration is approaching one million, with 314,689 commercial drones and 538,172 recreational drones registered at the time of writing. What was once considered a concept for the future has become commonplace. At the same time, drone-related security measures and technologies need to evolve as drones ultimately pose a variety of threats.

Examples include flying too close to restricted airspace, dropping contraband in prison yards, transporting drugs across borders, conducting prohibited surveillance and reconnaissance, and even dropping small bombs or explosives. According to the Cybersecurity & Infrastructure Security Agency, recognizing threats and implementing security practices that meet federal, state, and local regulatory requirements are critical steps in successfully addressing potential drone-related security incidents.

There are several methods of mitigating risk, starting with researching and implementing legally approved anti-drone technology and understanding the airspace around a facility or area. Autonomous detection systems like Walaris’ AirScout Sentry combine industrial cameras, powerful edge processing and proprietary computer vision and artificial intelligence (AI) algorithms for optical detection capabilities in critical applications.

Autonomous Threat Detection

Depending on the application requirements, the hardware-independent AirScout can employ low-cost cameras from Axis or Bosch, or use other off-the-shelf multispectral or infrared imaging cameras for longer-range missions. Walaris’ AirScout Sentry offers a complete detection, tracking and identification solution, while companies wishing to use other detection modalities such as radar can use AirScout Verify, a slew-to-cue tracking and identification software solution.

Each AirScout Sentry deployment of the system includes multiple detection cameras that continuously monitor an area of ​​interest. Detections from these video streams are relayed to a verification camera, which confirms the threat and warns operators or rejects false positive detections. If the AirScout Sentry software recognizes relevant data such as birds or a drone, it records the coordinates of the object. A pan-tilt-zoom algorithm causes a camera to zoom to the coordinates to further examine the object, while AI and range-to-focus control algorithms allow the system to find the target maintaining a clear image and to classify, identify and track it.

Once the system classifies an object as a drone, it provides the system operator with a warning along with full-motion video with the classification determination. The operator can then decide how to handle the situation. Different operators treat the information differently depending on the market they serve. In military or federal government operations, this often means engaging and weakening the target. At an airport, it might mean closing the runways, while a jail application might involve securing the yard and bringing in inmates.

A clean data collection pipeline

Classifying drones with a high level of confidence presents several challenges. According to Kyle Meloney, co-founder and CEO of Walaris in Atlanta, “They are small objects that move quickly and at a distance, clutter can exist in the area of ​​interest, and drones can fly over complex backgrounds.”

“Using cutting-edge AI algorithms, the system can perform optical classification in near real-time and with limited optical information, but getting to this point requires a lot of preparation when it comes to the data,” he said.

AirScout’s AI algorithms continuously train with a growing data set. Walaris’ data acquisition pipeline involves capturing images and cleaning that data to ensure all images are properly labeled so that the training process uses high-quality images to better train their models.

“Our training process is never-ending,” Meloney said. “We’re constantly adding clean, annotated images to the dataset and improving the algorithms over time, allowing the system to perform more effectively over time.”

Powerful edge processing for strong AI

The heart of the entire AirScout system is the software, which contains, among other things, the proprietary detection, classification and PTZ algorithms. Previous optical detection methods involved looking for movement and/or contrast in an image. This technique works well in simple scenarios, but Meloney says complications arise as the scene gets more complex.

“When a drone flies over a cloudy sky and there are birds in the scene, there’s a lot of movement and varying levels of contrast, and it can become difficult to spot the drone,” he said. “By processing the entire image in near real time and tagging each individual object to determine its relevance, we can reduce false positives and automate the ‘detect, track and identify’ response chain.”

Running complex AI algorithms requires powerful processing capabilities, so Walaris procured a custom-configured industrial PC from CoastIPC with an NVIDIA graphics processing unit (GPU) to run the AirScout software.

“Mission-critical applications require reliable, rugged PCs,” says Meloney. “Each PC shipped by CoastIPC has arrived on time and meets the exact requirements we specified to deliver and run flawlessly on every mission.”

eyes in the sky

There are several drone detection methods on the market, and no one option is necessarily the best possible solution. However, for those looking for a practical and comprehensive approach, systems utilizing cameras and AI software can provide comprehensive solutions for detecting, tracking and identifying drones in dynamic environments.

Leave a Comment