OmniPerception Limited exists to meet customer requirements in the fields of computer vision, facial biometrics and video analysis.
OmniPerception Limited exists to meet customer requirements in the fields of computer vision, facial biometrics and video analysis.
When OmniPerception was founded in the Spring of 2001, a total of more than 500 man-years of research and development had already been invested in the fundamental technology, intellectual property portfolio and expertise that lie at the centre of the company’s offering to customers as a provider of advanced image processing and computer vision solutions. OmniPerception’s principal founder, Professor Josef Kittler, is the long-time head of Surrey’s world-famous Centre for Vision, Speech and Signal Processing. He has dedicated the major portion of his professional life to the advancement of human 
knowledge and practical technological capability in this field.
The core technology that powers the OmniPerception product range was laid down by Professor Kittler and a committed team of researchers and innovators working under his leadership over more than 20 years at the highest level of scientific endeavour. In addition to his leadership role as head of one of the world’s largest and best know centres of excellence in vision, speech and signal processing, Josef Kittler is the author of more than 400 published articles in his chosen field. He was appointed a Distinguished Professor of the University of Surrey in 2004 and continues to develop and expand his own work and that of his department as well as supporting the company as Chairman of the OmniPerception Advisory Group. OmniPerception itself was founded by Professor Kittler and his colleagues Dr Charles Galambos and Dr Kieron Messer in the Spring of 2001.
At its inception, OmniPerception was a “spin-out” from the University of Surrey and, more importantly, the company remains a long-term technology partner of the University. OmniPerception has an exclusive intellectual property agreement with Surrey’s world-famous Centre for Vision, Speech and Signal Processing and is committed to continuous technological improvement and development. In this way, OmniPerception ensures the continuation of a fundamental research platform to underpin the company’s own in-house R&D; and the constant replenishment and refreshment of core OmniPerception technology, for the most effective deployment in the service of our customers.
OmniPerception’s Affinity™ Face Recognition technology provides uniquely fast and accurate performance in a wide range of face activated biometric applications including physical and logical access, e-Passports, ID Cards, time and attendance, prevention of fraud and identity theft and a range of applications in the field of criminal justice. (link to Affinity page)
Affinity™ is the brand name of the company’s proprietary face recognition technology and the suite of software built upon it. The company welcomes enquiries about its technology from customers and other interested parties and will be happy to provide more detailed information on its face recognition technology on request where appropriate. Only a relatively general set of notes is included here as an introduction to Affinity™.
In essence, Affinity™ face recognition technology is the next generation solution above and beyond the well-known “Fisher Face” technique of linear discriminant analysis. Building on the best features of the basic Fisher Face technique, and taking advantage of innovations pioneered at the Centre for Vision, Speech and Signal processing, Affinity™ technology generates more accurate and robust performance via optimized discrimination and specially enhanced “client specific” identification functionality.
Face recognition is often thought of as a single function – by both end-users and most face recognition technology vendors. However, face recognition is in fact a complex process with many stages and applications. Optimal approaches are required at each stage.OmniPerception Affinity™ face recognition segments and optimises each stage of the process and provides adaptive algorithms specifically designed to do each job optimally well:
Initial high quality and compliant facial image capture is essential to a future-proof success of face recognition. The ISO standard goes a long way towards ensuring high-quality image capture. However, certain issues still remain; the main one being lighting.Adverse lighting can transform a given face to such an extent that it can look like another person’s face and can therefore be mistaken both by human beings and computer vision systems. Therefore, for nearly all applications controlled lighting in some manner is required. Isotropic (uniform) lighting is ideally required to illuminate the face evenly so that the individual’s facial features are not transformed to the point where their face overlaps with another person’s face in ‘face-space’. OmniPerception’s patent-pending anti-spectral image capture devices go a long way to dealing with this problem. These are available in both natural light and night-sight versions.
In addition to the company’s world-famous Affinity™ face recognition, OmniPerception applies its core technology to a range of other advanced image processing applications. These fall into two broad categories; the tools necessary for the automatic identification, analysis and reporting of brand exposure and product placement in TV, film and other image communications media, through Magellan™ technology; and the application of specialized versions of this core technology to image analysis, object recognition and tracking in the fields of intelligence, reconnaissance and surveillance – branded Gama™.
Traditional object recognition techniques are only capable of achieving acceptable performance in complex analytical tasks by undertaking repeated searches using a very large number of templates. This is necessary to ensure the detection of objects appearing at different angles and different real or apparent sizes in the video stream. In contrast, OmniPerception's "Search Space Reduction" (SSR) technology achieves object recognition much faster and more efficiently.
This new approach to object searching and recognition makes innovative use of distinctive object features to allow accurate identification with much fewer search templates. Often a single template will suffice. OmniPerception’s complementary “Geometric Invariance” (GI) technology allows accurate analysis even when objects are presented within a huge range of random positioning, size and angle of presentation. OmniPerception's SSR and GI technologies are more accurate, more efficient and very much faster and more reliable than any other competing search techniques with additional advantages in greatly improved speed of processing.
The appearance of any object or "search target" in a video stream is greatly dependent on two key factors: The position of the object in space and the object's orientation with respect to the viewer (or the camera eye). Each of these two factors can be described according to 3 separate axes or dimensions. That is to say, the appearance of any one object requires the specification of 3 distinct parameters to describe it fully with respect to each of these two factors. To describe the object's appearance fully - in respect of both place and presentation - requires a total of 6 parameters for any one accurate description - or recognition - of the object. One of the most common techniques used for recognizing objects is template matching. However, template matching only deals with 2 of the 6 parameters involved in the task. It is possible to recognize an object, but only provided that the object is presented at or very close to the expected presentation. For multiple presentations, a search system based on template matching needs to use multiple individual search targets - and search processes. Even with the most advanced modern template matching techniques, successful object recognition from the use of a single "search target" requires a match to within no more than 15 degrees away from the expected presentation. With respect to angle of presentation alone, this means that the system in use needs to cover at least 12 possible angles in each of the 3 ways that the object can rotate. When scale is also taken into account, we have: 12 different scales, requiring 12-to-the-power-of-4 templates. As can easily be seen, this means that for fully accurate search coverage 20,736 templates would be needed for each object or search target under investigation; if conventional template matching techniques were being used. In contrast,
OmniPerception's approach, with search space reduction and geometric invariance, creates significant benefits in terms of both speed and accuracy. These features are of course highly relevant both to sports marketing and to the wider challenges of scene understanding in intelligence, reconnaissance and surveillance.In live tests in the UK and by the company’s teaming associates in the USA, on TV footage and on footage from surveillance and reconnaissance sources, OmniPerception Gama™ technology has, for instance, been successful in finding, recognising and differentiating
Within these categories, Gama is also capable of finding relatively small wordings or other distinguishing features on the sides (or other parts) of vehicles, pieces of equipment, cargo items and other objects, wherever distinctive markings can be found. In many cases, the distinguishing features may be characteristic of the objects themselves – e.g. a vehicle outline or shape of baggage item - and need not be separate markings. Pictographs or “search targets” need not be pre-selected. They can also be taken from observed footage – for instance the input from a cctv camera on one part of a site or premises – and then used to scan footage from elsewhere, both historic and live or future footage. For example, a vehicle that aroused suspicion can be “tagged” in this way and then monitored closely via the input from all site cameras (or indeed via cameras in other locations) until action is taken or the alert is stood down. This capability relates to vehicles and also of course to other objects and to people, provided that some sufficiently distinctive part of their appearance – eg their clothing, or what they are carrying - can be identified. Gama can find and identify such search targets even when their distinctive markings are only partly visible. The product has an advanced capability in this respect and can be fine tuned to meet specific customer needs. Where even greater detection reliability is required, the system can also search for a distinguishing feature in one or several parts, including quite small components, provided always that each is sufficiently distinctive.
Members of the OmniPerception Partners Programme and customers with special requirements are welcome to enquire further into the company’s technology and the OmniPerception continuous technology improvement programme.