Researchers at Universidad Carlos III of Madrid and the firms Abertis and Solusoft have developed an intelligent system that analyzes video surveillance camera images in real time, detects anomalous situations and alerts the nearest security agents in urgent situations, such as the presence of a vehicle moving in the wrong direction.
This platform, which combines artificial vision technology with geolocalization and ITC is part of DONDE, an R+D+i project that studies and validates mobility, localization and video processing technologies in order to offer new services to the public. Specifically, this system can be applied in the area of public and road safety, as the researchers explain. "It can automatically detect the presence of a driver moving in the wrong direction, recognize a danger, identify its position, indicate which security agent can handle the situation the fastest, and alert nearby drivers," declares Professor Antonio Berlanga, of UC3M's Grupo de Inteligencia Artificial Aplicada (GIAA- Applied Artificial Intelligence Group). In the area of emergency management, this system can accelerate decisions and shorten response time.
The main advantage of this platform is that it makes it possible to know exactly what is occurring and where it is occurring. The researchers add that it does not require an expensive, complex, new infrastructure to be built, because most cities already have surveillance cameras and other types of sensors that can be integrated into the system. "Furthermore, we include citizens in the concept of 360 degrees, because users do not just receive information, but they can generate it as well, using their smartphones", points out the Solusoft's Director of R+D+i, Sergio Alcalde. For example, with the implementation of this platform, a person who notices a dangerous situation caused by a specific object (a sewer without a grate, a cornice that may fall off, etc.) can photograph that object and situate it using a mobile phone's GPS, and then send the photo to the corresponding maintenance services so that they can intervene as soon as possible.
The team has developed a device to demonstrate different scenarios in the Abertis Smart Zone, which is in the company's Barcelona facility. Thanks to these scenarios, they have been able to prove the technical viability of this "intelligent" system, which is able to comprehend what is happening. Until now, visualization systems depended on an operator who had to be constantly looking at what the cameras were transmitting in order to detect any safety-related incidents. In contrast, this platform's sensors (the cameras) are equipped with the capability of processing what they see and reporting on what, when and where an incident has occurred. "Considering the number of images that control centers must visualize, it is good to have a tool that brings attention to images of those points where the intelligent system has already detected that something dangerous may be happening," explains Professor Berlanga.
The GIAA is formed by an interdisciplinary team of researchers from UC3M. Their work is oriented toward solving engineering problems using the latest in artificial intelligence technology: automated learning, evolutionary computation, data analysis, multiobjective optimization, fuzzy systems and intelligent agents. The group has a great deal of experience advising companies and developing made-to-measure solutions for data fusion, prediction, optimization, and signal and imaging processing.
The DONDE project, which received funding as part of the Ministry of Industry, Tourism and Commerce's Strategic Action for Telecommunications and the Information Society (Acción Estratégica de Telecomunicaciones y Sociedad de la Información) of the Avanza 2 Plan, was carried out by a consortium coordinated by Abertis Telecom with participation by the University's GIAA and Solusoft, a company located in the Parque Científico de Leganés Tecnológico (Leganés Technology and Science Park), just one hundred meters from UC3M's Science Park, a meeting place for innovation that the University participates in; it is a productive business area that facilitates collaboration between UC3M researchers and local firms. The consortium's experience has been very positive, according to its members, as it has led the three entities to design a common platform of services, based on perceived needs, which both private companies and public administrations will be able to use to provide the public with new services that will improve public and road safety.
Explore further: Pedestrian crossings could be monitored
More information: Parque Científico UC3M (UC3M Science Park): www.uc3m.es/pcyt
DONDE project website: www.proyectodonde.es