Technology - Video Analytics
Advanced Image Processing / Video Analytics
OmniPerception is committed to the on-going development and deployment of its core technologies in a range of advanced image processing applications where truly transformational technology is needed to meet customers’ requirements in mission-critical areas.
Designed to give customers the new higher levels of performance and reliability they need, the company’s advanced image processing solutions fall into two broad categories:
Magellan™ Technology
OmniPerception’s Magellan™ technology drives a range of Magellan™ products that detect, analyse and record the occurrence of logos, brands and other similarly distinctive two dimensional images in television footage, video and other image based media. The software detects and highlights each logo or other “search target”, adding a label to identify it for the human user and including a comprehensive set of associated information required by the customer (eg. position, size and degree of occlusion). Magellan™ from OmniPerception is the world’s number one choice for solutions in this field and is currently processing more than 20 terabytes of image data every month for satisfied customers on four continents.
Magellan™ is capable of detecting an image very accurately and very fast even when only a small part is visible. It can also detect images that are somewhat blurred and images presented in oblique and even quite severely distorted presentations. This includes the detection of insignias and slogans on sports clothing and vehicles as well as on the pitch, track-side billboards, race circuit bridges etc.
The way in which Magellan™ operates can be tuned very finely to fit the needs of an individual sponsor or brand owner, in terms of acceptable presence on screen, size, occlusion, position, and degree of blur or “picture quality”. The software collects all the necessary information in one pass, but the results can be re-tuned and re-presented as many times as necessary in response to specific customer enquiries about specific aspects of the footage – almost instantaneously, without the need to re-run any of the footage.
Some of the key distinctions between OmniPerception’s unique technology and more traditional and less effective approaches to the challenges in this field are covered at more length in the Gama™ section below and in the separate downloadable PDF available to customers and partners seeking more background information on OmniPerception’s unique capabilities in this field.
Gama™ Video Analytic and Scene Awareness Technologies – The fundamental advantage
Practical cost-effective 'intelligent video' installations need to be equipped with powerful core technology enabling advanced image detection and recognition at very high speed. They also need to be capable of operating on reasonably sized and priced hardware platforms. Both requirements present a huge challenge for traditional object recognition techniques. Traditional technologies are rarely capable of achieving acceptable performance without undertaking repeated searches using a large number of templates, with inevitable disadvantages in terms of both processing times and capital cost. They also tend to suffer horribly from high levels of 'false alarm'.
The key problem is that with traditional approaches, multiple templates are necessary to ensure the detection of objects appearing at different angles and different sizes in the video stream. This creates a need for very large hardware installations. Even then, traditional technology is rapidly overpowered when the video analytic task becomes at all complex. In contrast, OmniPerception's 'Search Space Reduction' (SSR) technology achieves object recognition much faster and more efficiently. With SSR, often a single template will suffice, delivering huge savings in processing time and hardware cost, as well as significant reductions in false alarm rates.
OmniPerception’s complementary 'Geometric Invariance' (GI) technology allows accurate analysis even when objects are presented within a huge range of random positioning, sizes and angles of presentation. This makes solutions based on OmniPerception technology more accurate, more cost-effective, faster and more reliable, as compared with more traditional video analytic techniques available from competitive suppliers.
A supplementary downloadable PDF is available to customers and partners seeking more background information on OmniPerception’s unique capabilities in this field.