London-based systems integrator NMR is leading a consortium of European developers that has been awarded €1m (£840,000) of EU funding to build and commercialise content recognition software.
The software will automatically analyse and extract time-stamped descriptive and technical metadata from live broadcastquality video content, as well as archive content.
The potential use cases listed by the consortium include real-time content recognition and technical monitoring of live broadcast feeds and identifying individuals, objects or locations in a large media archive.
The consortium working on the Real-time Content, Analysis and Processing (ReCap) platform also includes ToolsOnAir, Nablet and Joanneum Research.
They will combine existing technology into a “scalable and open” system that will deliver analysis of video and audio content that includes logo, face detection and content duplication.
The consortium also wants the software to automatically detect quality control issues such as lost or frozen video frames, dropouts, visible macroblocking and estimation of sharpness and noise.
It could potentially include automatic quality improvement, interpolation of image regions or entire frames, denoising and block artefact suppression.
The ReCap product will support on-premise, data centre and cloud deployment models.
“The ever-increasing volume of video content, combined with a lack of time and human resources to view, analyse and process it, makes the timing perfect for the consortium to bring project ReCAP to market,” said NMR chief executive Neil Anderson.
“This EU grant has also enabled NMR to employ additional full time software engineers, including lead developer and ReCAP product owner David Moss, who previously contributed substantially toward the development of the Zonza production management platform, while working at Hogarth Worldwide, part of WPP.”
The ReCap project is funded by a €1m grant from EU Research and Innovation Programme Horizon 2020.