IP protection




Please provide a short description of the state-of-the-art and/or current trends in the field? How does the result fit into it?
Keywords: security system, surveillance, cognitive system

The increasing number of surveillance cameras in a lot of different contexts underlines the need for systems able to autonomously identify abnormal behavior within the field of view of a surveillance camera in real time – enabling security personnel to monitor exponentially more video in a more effective and faster manner. To reach that objective some learning mechanisms are needed to allow decisions about events different from normality. Cognitive systems mimic human brain processes: learn from what it observes, remember activity patterns and adjust to changes in the environment. Applying cognitive techniques to surveillance is currently offered by some products on the market. Some examples are:

AISight ( http://brslabs.com/index.php?id=74 )is a Behavioral Analytic software product built, tested and proven in the field to enhance video surveillance by providing automated awareness of abnormal behavior. Taking visual input from either a live camera or recorded video, AISight autonomously learns what activities and behaviors normally occur in an environment or scene. When AISight observes anomalous behavior, the software generates an alert to notify appropriate security personnel in real time. The system can be managed and monitored through a user-friendly desktop and browser-based application, or can easily integrate into existing video management or command and control systems. AISight installation is as simple as installing the software and linking the camera feeds. AISight begins observing and learning the minute the system is turned on. No coding of rules or triplines is required. No areas within the camera field of view are masked out. Because AISight observes, learns and alerts on behaviors, instead of traditional video motion detection, AISight performs very well across complex environments including dense moving foliage, moving crowds, and myriad weather conditions.

VisionKit™ ( https://www.cra.com/commercial-solutions/computer-vision-components.asp ) computer vision systems enable computers to automate repetitive visual tasks, freeing human experts to concentrate on higher-value activities. VisionKit™, Charles River Analytics’ library of computer vision components, allows developers to rapidly prototype vision systems using our advanced technology. The components are written in C++, and are intended for real-time performance, while still offering a developer-friendly algorithm prototyping environment.




What is the problem/need/knowledge gap that the research result is responding to?  How was it addressed before?
The research result is something quite general an it could be applicable to many different tasks. At a higher level of abstraction, the system shall help humans in some surveillance task where a constant level of attention is required even if a low rate of relevant events is likely to happen. In that case, repetitive situations may cause a loss of attention potentially leading to dangerous situations. The machine vision could certainly make the first level of surveillance drawing attention only to those events requiring human intervention.

The project aims to develop a system that is capable of gathering, fusing, interpreting and producing information about parts of a risky industrial environment through recognition and description of objects and activities. The proposed behavior analysis surveillance system is able to monitor human activities in high-risk areas and to deduce about the normality of the situations that individuals are getting involved. The representation and interpretation of behaviors is possible through a harmonic combination of low-level features (geometric), which will be extracted, from visual data and high-level knowledge data (semantics). The system has relevant applications for industries that construct different products in the same assembly line, since the assembly procedure differs from product to product. Apart from the object recognition task, the range data can be further processed to perform visual inspection of the products.

The approach followed by the R&D result is to enhance the performance of the vision system by using a feedback channel between the central video processing unit and the remote cameras. In that case the cameras “collaborate” with the algorithms to provide the best obtainable result in a certain operating condition.

In common system, cameras and algorithms work separately and the algorithms have the responsibility to gather the best of the information contained in the video, without the possibility to tune the video cameras in a way the algorithms give the best performance. Furthermore, autonomous decisions are not often made by the monitoring system and the human operator has to be constantly focused on detecting potentially relevant events.




What is the potential for further research?
The potential is represented by the discovery of new cases by applying the system in  the real environment and, thus, by the needs, coming from the field, to solve new problems or to improve current algorithms and techniques.




What is the proposed method of IPR-protection? (patent, license, trademark etc.)
The patent could be a possibility for a complete “machine” to be considered as a product. At the moment, the product will be based on some commercial off the shelf components like a a processing unit (a PC), the cameras and the scanners and some algorithms. The core of the system is in the algorithms and in their customization and tuning to the special case in which they have to be applied. Whenever a commitment from a customer will require the implementation of the system in a production line, the patent will be considered to protect a well defined setup of hardware and software.





What are the steps that need to be taken in order to secure the IPR-protection? What is the cost of IPR-protection?
A first view of patents with the keywords, security system, surveillance and  cognitive system shows few possible relevant results. One of them is the following:


Actually, the system is a complex of many components, specially on the algorithm side, it is advisable to consider existing patent relevance case by case.

The patent costs in Italy are around 2500 € (up front) + increasing annual fees from 60€ up to 650€ in the 20 year period. The estimated cost for the PCT application in the main countries is in the range of tens of k€.




What is you overall assessment of the scientific maturity of the research result?
The scientific contributions used in the project are coming from a research work that has been tested in a controlled environment. To reach a deliverable product, a customization of the current work to the specific application is needed. After the customization a tuning of main parameters has to be performed on the field to adapt the general algorithms to the real boundary conditions. Thus, the remaining task are in the field of engineering and validation, this is why it can be stated that from the scientific viewpoint the result is mature.




Please put X as appropriate. 1 2 3 4 5
Scientific maturity       X  
Synergies     X    
State-of-the-art/innovation     X    
IPR-potential     X    


Bookmark the permalink. Follow any comments here with the RSS feed for this post. Post a comment or leave a trackback: Trackback URL.

Post a Comment

You must be logged in to post a comment.

Request a proposal

Valorisation Plan Authors

Related Documents

There in no related documents

Visit the other applications of the INTERVALUE Platform: R&D Repository | IP Agreements

© 2009-2010 INTERVALUE Project