European researchers working in the MOBVIS project have successfully developed a system that can attach hyperlinks to real-world urban features. Users simply take a picture of a streetscape with their mobile phone and the MOBVIS system will do the rest.
It will accurately identify the features, such as buildings, monuments or even street furniture. It will then render relevant information by applying an icon, within the picture on your phone, to the feature. Simply click on the icon to get the information directly. MOBVIS can even identify logos and banners.
The potential range of applications is vast. Click the icon to get details about what bus routes are served by a particular stop, and at what time. Quickly scan today’s specials on a restaurant menu or the best shopping bargains on the street.
Tourism is an obvious application. Instead of getting a guidebook that people read to discover where to go, take a picture and find out the history and culture of where you are. See a building that looks interesting? Take a picture and find out more about it.
Real estate, too, could be a fertile application. If house hunters see an area they like, they can take a picture of the streetscape to find what is on offer. Then users would simply select the icons to get further information and, ultimately, perhaps book a viewing. These types of application are just the tip of the iceberg.
“Really, we do not know all the applications that could potentially be developed from a technology like this. We can easily imagine some of the obvious ones, but after that, who knows?” admits Lucas Paletta, coordinator of the MOBVIS project.
It is like SMS. When it first came out, it was merely a simple way to communicate between users and the network. But cash-strapped callers soon realised its potential for low-cost communication, and then came ring-tone distribution, wallpapers and any number of other commercial applications.
MOBVIS’ system can recognise urban semantic information, like pedestrian crossings and traffic signs, and it can extract from that meaningful information about the context of the image. The system knows, for example, that the user is walking through pedestrian zones with shops.
The context enables the system to localise areas within the image where specific features will occur. This extrapolation of detail from the context can help the system to recognise and interpret its surroundings, and direct the user to specific points within the city.
A further service developed in MOBVIS is mobile multimodal positioning, which simply means using information from a number of sensors, like GPS, inertial sensors, camera, step counter and so on. It can be particularly useful in areas where GPS signal quality is poor, such as urban canyons, and areas with a high degree of signal reflection.
At this point, the MOBVIS platform morphs from an interactive guide to a positioning system. This aspect of the system opens the way to geo-services, such as giving users directions.
MOBVIS partner Tele Atlas provided geo-referenced imagery to the project. Tele Atlas, a world leader in providing digital maps for various kinds of geo-services, continuously collects geo-referenced imagery on road and urban infrastructure. The company has a fleet of more than 50 mobile mapping vans worldwide, constantly collecting this data.
It was reference photographs taken by Tele Atlas mapping vans that seeded the MOBVIS database with its initial cityscapes, namely the inner cities of Graz (Austria), Ljubljana (Slovenia) and Darmstadt (Germany). Geo-referenced imagery from the streetscapes is now available as a resource, online, for the scientific community.
The system, allied to geo-services like those offered by Tele Atlas, could eventually produce a ‘seeing-eye’ mobile phone application, for the visually impaired, though this is currently a distant prospect. Such a system would need enormous testing and validation before it could go live, but it is a testament to the work of MOBVIS that the concept can even be considered.
Picture-driven search engine
All these factors mean that MOBVIS results have a very high potential for creating revolutionary applications.
The MOBVIS interface ultimately works like a picture-driven search engine about things that just surround the user. The picture is your search query, and the system matches the picture and the features it contains with results in its database.
Users then select the ‘search hits’ by picking the hyperlink on the object of interest. The particular contribution by MOBVIS is to enable free exploration of urban space instead of needing a guide to point to specific opportunities for interaction.
It really is a paradigm changing innovation, and one that will drive any number of new and emerging applications over the months and years to come.
This is part two of a three-part feature on MOBVIS by ICT Results.
The MOBVIS project received funding from the Future Emerging Technologies FET Open strategic objective within the ICT strand of the Sixth Framework Programme for research.
Christian Nielsen | alfa
21.08.2017 | Albert-Ludwigs-Universität Freiburg im Breisgau
AI implications: Engineer's model lays groundwork for machine-learning device
18.08.2017 | Washington University in St. Louis
16.08.2017 | Event News
04.08.2017 | Event News
26.07.2017 | Event News
21.08.2017 | Medical Engineering
21.08.2017 | Materials Sciences
21.08.2017 | Life Sciences