The Magellan platform is the technology behind our CloudVis solution for identifying a logo, graphic or shape in a section of video.
Our video analytics software can measure a brand's exposure on screen, or it can be used to support investigations that involve forensic analysis of CCTV footage.
Key suspects or objects can be isolated and tracked through hours of footage simply by pinpointing a brand, logo or shape on their clothes.
The system is capable of analysing video data from a broad range of inputs, providing an efficient approach to an otherwise effort-intensive activity.
Unlike human operators, Magellan remains alert and attentive – even in the face of large volumes of CCTV footage or over long periods – and can reliably automate surveillance tasks that would otherwise place significant demands on staff.
It also reduces the elapsed time in searching through critical video. This allows operators to identify vehicles or subjects of interest quickly, using markers such as logos, icons or numbers.
COVERS ALL ANGLES
Provides detection of object from every perspective.
ACCURATE AT HIGH SPEED
Performs multiple brand and logo searches in a single measurement.
Capable of functioning at 25 frames per second and will work with most video formats.
Measures how long a searched item remains on screen.
Our video analytics software reports reliably on the presence of logos in live or pre-recorded video.
The technology will accurately calculate the number of times logos appear, enabling the production of comprehensive reports or analysis.
The Magellan tool is able to process large volumes of recorded CCTV footage in order to identify instances of objects of interest. It uses a template model, specified by an operator, to input objects into a search.
It is also able to identify any objects of interest appearing in a real-time video stream – even where the object may be inverted or partially obscured.
Automates and enhances surveillance tasks, using reliable real-time or post-event object detection
Ability to detect objects when partially obscured or presented at oblique angles or on irregular surfaces
Processes large volumes of archive CCTV footage to track subjects of interest, using visual identifiers
Ability to build up multivariate templates, criteria and filters to refine analysis of archived video footage
Simultaneous identification of several objects within footage – with customisable output reporting
Video ingest and template definition across multiple workstations, with simple scale-up of processing
Accurate, reliable and repeatable processing around the clock for increased productivity and cost savings