ChyronHego has introduced TRACAB Gen5 optical sports tracking system, featuring improvements in tracking data quality and accuracy, driven by redesigned tracking algorithms, an array of camera angles, and AI features for player, number, and colour recognition.
TRACAB Gen5 uses a distributed system of cameras installed around the field of play and advanced image processing technology, to capture and deliver real-time tracking data on the movements of each player, referees, and the ball. TRACAB has been installed in over 300 stadia and is currently used to capture live tracking data for more than 4,500 football/soccer and baseball games each year.
"We've overhauled and improved our AI-based image detection tracking algorithms from the ground up to ensure maximum accuracy and the lowest latency. We're combining these advanced algorithms with multiple high-resolution camera angles, which means we can deliver the highest quality data feed on the market in real-time, and provide an greater value for leagues, federations and teams, betting companies, broadcasters, and OTT rights holders producing virtual graphics and enhancements for better fan engagement," said Rickard Öhrn, president - Sports, at ChyronHego.
TRACAB Gen5 includes a distributed camera architecture in which cameras can be deployed on both sides of the field and behind each goal to capture action from four angles. Gen5 is able to track every object on the field at a much higher resolution. The accuracy of player identities is further enhanced through TRACAB Gen5's AI capabilities, which enable the system to recognise and distinguish player numbers and jersey colours from any angle.
TRACAB Gen5 is able to track objects on the field with an average accuracy of 7cm (relative to the centre of mass of a player) and 100% tracking completeness. Coupled with a real-time latency of below 300 milliseconds, these are the highest performance results ever achieved by any sports tracking system in the world. ChyronHego is expanding the utility of TRACAB's convolutional neural network algorithms even further to deliver limb or skeletal tracking. This is done by identifying players' key body parts, such as heads, shoulders, arms, hips, knees, heels, and toes, to enable the creation of three-dimensional skeletal movements of players. The resulting multidimensional data, accurate enough to define the position of the tip of a striker's foot, will be fed automatically into VAR (Video Assistant Referee) operations. This will allow offside calling decisions to be powered by instantaneous and automatic identification of situations.