Our computer vision technology centers on automated image and video analysis, which can be applied to any domain where images are a vital part of business. Our team has expertise in extracting data from both video and still images. We seek to continually advance the field of computer vision through research and development (R&D) and through collaborative projects that build on our open-source software platform, the Kitware Image and Video Exploitation and Retrieval (KWIVER) toolkit.
Our customers include government entities such as the Defense Advanced Research Project Agency (DARPA) and Air Force Research Laboratory (AFRL). We have worked with these entities to deploy an operational Wide Area Motion Imagery (WAMI) tracking system for Intelligence, Surveillance, and Reconnaissance (ISR) in theater as well as content-based image and video retrieval capabilities. We have also worked with private companies such as Lockheed Martin and Raytheon through consulting services and collaborative projects.
Areas of Focus
Social Multimedia Analysis
Our large-scale multimedia analysis tools can automatically understand content from millions of videos and images shared yearly on social media. This analysis incorporates our expertise in object detection, scene understanding, and event recognition to provide accurate content-based search capabilities and robust privacy protection.
Object Detection and Recognition
Our object detection and recognition tools provide the abilities to identify and track objects in WAMI and full motion video (FMV). These tools can handle difficult scenarios such as low resolution, low contrast, moving cameras, occlusions, vehicle changes, and situations in which there are a large number of targets.
Event and Activity Recognition
Our event and activity recognition tools analyze events, behaviors, activities, and transactions in a scene. For example, our tools can characterize, model, and detect events such as people carrying objects, people pushing/pulling objects, people running, vehicle starting/stopping, and vehicle making U-turns. Our methods work in both WAMI and FMV, as resolution permits.
Our scene understanding tools use machine learning techniques to provide complex functional object recognition in WAMI video. Through defining objects by their behavior rather than their appearance, these tools can differentiate uses of similar objects. For example, they can distinguish a car used for police activities to one used for pizza delivery. The tools can also detect complex threat patterns such as improvised explosive device (IED) emplacement.