Computer Vision
Discover cutting-edge solutions from 5 leading global suppliers
Vision4ce has released a next-generation tracking feature for its CHARM video target tracking hardware products. The incorporation of machine learning and AI into the trackers’ image processing provides improved performance in complex scenarios, including automated acquisition and reacquisition of targets.
The software upgrade ensures reliable targeting of small agile targets such as single drones and swarms and autonomous surface vessels, without the need for intervention by the operator when a target becomes obscured or changes perception angle. This capability also extends to vehicles and people in crowded ground scenes as well as aircraft in the presence of cloud clutter.
This new feature, for the CHARM 100 and CHARM 100 NX platforms, exploits AI-based methods to maintain tracks in complex scenes. This provides increased probability of reacquisition and reduced frequency of triggering by false targets due to improved bounding of the target. Automatic target reacquisition removes the requirement for manual reacquisition following loss of a track, thus reducing the burden on the operator and simplifying the target acquisition process.
John Thornton, Director of Image Processing at Vision4ce, commented: “This is a significant capability enhancement for our customers. The CHARM tracker is now even more resilient, capable of maintaining robust tracks even with challenging targets, and provides enhanced adaptability to rapid changes of the target, augmenting an already impressive range of products.”
“This software upgrade showcases the commitment of the Vision4ce brand to continuously develop capabilities that both lengthen the life and augment the performance of customers’ products – innovations that ultimately meet ever-evolving requirements.”
