Stottler Henke to Develop an Adaptive Threat ID System for Navy Pilots
Advanced software will combine computer vision, statistical analysis, and artificial intelligence to identify threats to aircraft quickly and accurately
San Mateo, CA, March 18, 2016 --(PR.com)-- Stottler Henke Associates, Inc. (www.stottlerhenke.com) today announced the award of a contract with the U.S. Navy to develop the Adaptive Target Threat Identification System (ATTIS), an advanced software system that will analyze infrared and thermal imaging video data to identify threats to aircraft automatically. ATTIS will dynamically combine computer vision, statistical analysis, and artificial intelligence techniques to identify threats accurately at a wide range of distances.
Today’s Navy aircraft such as the F/A-18 strike fighter and the F-35 stealth multi-role fighter carry numerous infrared and thermal imaging devices and sensors to provide tactical information to pilots. These systems detect and track objects, but they cannot identify them. In order to identify objects and determine threat levels, pilots must observe the data manually. This process is time-consuming and takes the pilot’s attention away from other critical tasks. Identifying objects and determining threat levels earlier and at a greater distance would provide pilots with more time to process the information and determine the best course of action. However, at far distances, detected objects are only a couple of pixels in size, down to less than a pixel in size, so manual threat identification is extremely difficult.
ATTIS will identify objects accurately, even at large distances, by dynamically adjusting its processing method depending on the input data. It will combine traditional computer vision techniques for object classification with statistical analysis on pixel data when the pixel size of the object is too small to analyze using traditional vision algorithms. When the detected object is close (i.e., several pixels or more in size), the image of the object will be sharpened by combining several frames to gain sub-pixel resolution, and the image will then processed by the object classifier module using computer vision techniques. The statistical analysis module will estimate the object’s location, speed, and size from object detection and tracking data. The behavior analysis module will also input object detection and tracking data to classify the object’s maneuvers and actions using Stottler Henke’s SimBionic(R) intelligent agent toolkit. SimBionic runs many hierarchical finite state machines in parallel to detect meaningful temporal patterns of events and state conditions that indicate threat behaviors and intentions.
ATTIS will intelligently combine the output of the object classifier and statistical and behavioral analysis modules to perform the final threat level identification. Low-level feature extraction and high-level behavioral pattern and logical analysis will detect distinctive characteristics of each target. Object characteristics such as heat signatures, movement over time (including minimum and maximum speeds and altitudes), and size will be correlated with an existing database of aircraft and missile types to determine the object’s type, classification, and trajectory, which, taken together, determine the threat level.
"The Adaptive Target Threat Identification System is unique in its ability to combine diverse computational and reasoning methods to extract the most information possible from the infrared sensor data," says Richard Stottler, the project’s principal investigator. "The ability to accurately identify threats at greater distances will enable U.S. Navy pilots to see farther and act sooner," continues Stottler.
During this contract, Stottler Henke will demonstrate the effectiveness of the combined technologies.
Founded in 1988, Stottler Henke Associates, Inc. applies artificial intelligence and other advanced software technologies to solve problems that defy solution using traditional approaches. The company delivers intelligent software solutions for education and training, planning and scheduling, knowledge management and discovery, decision support, and software development. In 2012, Stottler Henke, in a White House ceremony, was awarded the prestigious Tibbetts award, which honors small businesses for outstanding technical achievements and innovativeness. US Government agencies have designated ten Stottler Henke systems as Small Business Innovation Research (SBIR) success stories. Four Stottler Henke systems have been included in Spinoff, NASA's showcase of successful spinoff technologies. Stottler Henke was the subject of a NASA Hallmarks of Success video profile for its work developing and later commercializing advanced scheduling and training software systems. Stottler Henke received a Brandon Hall Excellence in Learning award for innovative technology. Stottler Henke was named one of the "Top 100" companies making a significant impact on the military training industry by Military Training Technology magazine for 2014 and ten previous years. Stottler Henke has received a Blue Ribbon from Military Training Technology magazine, recognizing it as a company that leads the industry in innovation. Email: info@stottlerhenke.com. Web:
http://www.stottlerhenke.com.
Today’s Navy aircraft such as the F/A-18 strike fighter and the F-35 stealth multi-role fighter carry numerous infrared and thermal imaging devices and sensors to provide tactical information to pilots. These systems detect and track objects, but they cannot identify them. In order to identify objects and determine threat levels, pilots must observe the data manually. This process is time-consuming and takes the pilot’s attention away from other critical tasks. Identifying objects and determining threat levels earlier and at a greater distance would provide pilots with more time to process the information and determine the best course of action. However, at far distances, detected objects are only a couple of pixels in size, down to less than a pixel in size, so manual threat identification is extremely difficult.
ATTIS will identify objects accurately, even at large distances, by dynamically adjusting its processing method depending on the input data. It will combine traditional computer vision techniques for object classification with statistical analysis on pixel data when the pixel size of the object is too small to analyze using traditional vision algorithms. When the detected object is close (i.e., several pixels or more in size), the image of the object will be sharpened by combining several frames to gain sub-pixel resolution, and the image will then processed by the object classifier module using computer vision techniques. The statistical analysis module will estimate the object’s location, speed, and size from object detection and tracking data. The behavior analysis module will also input object detection and tracking data to classify the object’s maneuvers and actions using Stottler Henke’s SimBionic(R) intelligent agent toolkit. SimBionic runs many hierarchical finite state machines in parallel to detect meaningful temporal patterns of events and state conditions that indicate threat behaviors and intentions.
ATTIS will intelligently combine the output of the object classifier and statistical and behavioral analysis modules to perform the final threat level identification. Low-level feature extraction and high-level behavioral pattern and logical analysis will detect distinctive characteristics of each target. Object characteristics such as heat signatures, movement over time (including minimum and maximum speeds and altitudes), and size will be correlated with an existing database of aircraft and missile types to determine the object’s type, classification, and trajectory, which, taken together, determine the threat level.
"The Adaptive Target Threat Identification System is unique in its ability to combine diverse computational and reasoning methods to extract the most information possible from the infrared sensor data," says Richard Stottler, the project’s principal investigator. "The ability to accurately identify threats at greater distances will enable U.S. Navy pilots to see farther and act sooner," continues Stottler.
During this contract, Stottler Henke will demonstrate the effectiveness of the combined technologies.
Founded in 1988, Stottler Henke Associates, Inc. applies artificial intelligence and other advanced software technologies to solve problems that defy solution using traditional approaches. The company delivers intelligent software solutions for education and training, planning and scheduling, knowledge management and discovery, decision support, and software development. In 2012, Stottler Henke, in a White House ceremony, was awarded the prestigious Tibbetts award, which honors small businesses for outstanding technical achievements and innovativeness. US Government agencies have designated ten Stottler Henke systems as Small Business Innovation Research (SBIR) success stories. Four Stottler Henke systems have been included in Spinoff, NASA's showcase of successful spinoff technologies. Stottler Henke was the subject of a NASA Hallmarks of Success video profile for its work developing and later commercializing advanced scheduling and training software systems. Stottler Henke received a Brandon Hall Excellence in Learning award for innovative technology. Stottler Henke was named one of the "Top 100" companies making a significant impact on the military training industry by Military Training Technology magazine for 2014 and ten previous years. Stottler Henke has received a Blue Ribbon from Military Training Technology magazine, recognizing it as a company that leads the industry in innovation. Email: info@stottlerhenke.com. Web:
http://www.stottlerhenke.com.
Contact
Stottler Henke Associates, Inc.
Jim Ong
(650) 931-2710
www.stottlerhenke.com
Jim Ong
(650) 931-2710
www.stottlerhenke.com
Contact
Categories