24 Aug 2010
By 2013 all municipalities in Germany have to introduce a double book-keeping. For the opening balance sheet of this system an assets and liability collection and evaluation of all municipalities is necessary. This includes the condition of the roads. Whereas some municipalities send employees with measuring sticks and metering wheels on the road, others, especially larger cities assign those tasks to surveyor’s offices which partially use imaging high tech for data acquisition.
This is the reason why you do not only find cars taking images for Google street view on the roads right now, but also vehicles from engineering companies equipped with complex roof installations. Triggered depending on the location, views of the road from different angles at a few meters distance are taken using modern camera technology. The determination of the position is done with an accuracy of 0.5m using a combination of GPS data and the information from a position measuring system and an inertial platform that allows a highly precise measurement of the angle and the acceleration on all three axes. Several cameras deliver image information from the drivers view, document the road condition, the roadway width, road marking, marginal strips, crash barriers and pedestrian tracks. The PC systems installed inside the vehicle are optimized for long-term sequence recording of several camera channels. Image analysis is done afterwards in the office, depending on the special requirements from the customer.
The Technical University of Dresden made there first tests in this field in 2002 with a simple system for front image acquisition, based on two Marlin Firewire colour cameras from Allied Vision Technologies, Dr. Dirk Ebersbach describes the first attempts in this field. He and his team from TU Dresden produced the first prototype of such a special vehicle back then. His partner in developing the system was the engineering consultancy Ulf Neubert, who have for more than 20 years been sales partner of STEMMER IMAGING in the eastern part of Germany.
Experience gained from this first model that ran for about 200,000 km with only one breakdown, was used by Ebersbach and Neubert several years later using Gigabit Ethernet technology. The subsequent development had two additional rear cameras and was able to document complete road conditions (potholes, breaks, ruts). At this time the TM-2040 CCD-camera from JAI with its fast dual tap mode, enabled us for the first time to acquire images at the required speed in a distance of 1m, Ebersbach remembers. We might have been the very first users of these cameras, as our models had the serial numbers 1 and 2, These cameras were complemented by two front cameras, model JAI TMC-4200.
In the meantime the knowledge gained at the TU Dresden was driven further by several companies. As an example, the proven rear camera-system was installed to a road condition inspection vehicle of the company Lehmann + Partner GmbH, where Mr. Ebersbach is working now. Using these cameras and a special illumination system, thousands of road kilometres in Germany are captured.
An important precondition for practical use in this field is the authorization for image acquisition by the German Federal Road Research Institute. “Our solution passed all tests without problems. The cameras worked reliably and provided a very good resolution states Ebersbach.
A challenge in any respect was a different project in 2009 where the single image technology had to be adapted to twelve cameras on an inspection vehicle of the company eagle eye technologies GmbH. The camera system also had to be adapted to the highly precise in-house navigation solution of this service provider. Besides handling of incoming data, PC technology and power supply had to be adapted to this challenge. We first thought about using CameraLink technology, but then decided to use Gigabit Ethernet. This decision has proven to be a blessing, as we have experienced no problems so far during operations, Dr. Ebersbach describes the course of this project.
According to Dr. Ebersbach a close cooperation with Ulf Neubert and STMMER IMAGING as competent imaging partners was very important for this success. “When we started developing our third system, Gigabit Ethernet technology was not yet widely used in imaging. We could take much benefit from the informed advice and long-term experience of STEMMER IMAGING in all technical questions.
The use of weatherproof and fully integrated camera protection housings for the roof of the vehicle proved to be a big advantage. These IP66 camera heads can be taken off individually and be mounted onto different vehicles according to demand. This advantage is especially important for use at eagle eye technologies, as their fleet does not only contain several inspection vehicles for highways, federal roads and municipal roads, but also includes an all terrain narrow gauge vehicle for pedestrian tracks, forest roads or dirt tracks.
Road inspection vehicles as described in this article enable acquisition of image sequences at a speed of up to 130 km/h. The incoming data volume is immense and required an elaborate concept for storing of image data. For this reason only sensor raw data is stored. A conversion to RGB for video reasons is done afterwards if required.
The images are stored including geo referencing. With the aid of a special photogrammetric software coordinates of the different objects can be determined during a later analysis. Traditionally these objects had to be registered by a surveyor individually on-site, requiring a lot of time. With the help of the software solution developed by Ebersbach and his colleagues, the effort is reduced to a few minutes.
The result enables direct visualization in a geographic information system such as Google Earth. Data can be transferred quickly and effectively and is often the basis for political decisions concerning further development of the infrastructure.
According to Dr. Ebersbach the current inspection vehicle is a good basis for further ideas. He could imagine using similar systems also for tunnels or railway tracks, or for route planning of heavy duty transports, according to road network data. Another idea would be the use of this data for a more handicapped accessible design of traffic infrastructure.
The solutions developed in this area of application are also interesting for another reason. A large percentage of the imaging components used there, where mostly used in industrial applications. The objects to be inspected are moving in front of the imaging system and not the imaging system itself is moving. However, the trend towards non-industrial imaging applications will increase a fact that is also stated by the current market survey from VDMA (Verband Deutscher Maschinen- und Anlagenbau). According to this survey the percentage of turnover from non-industrial applications raised from 10.0% in 2008 to 16.2% in 2009. VDMA assumes that the importance of on-industrial imaging applications will increase further.
The Departement Gestaltung von Straßenverkehrsanlagen (Chair of Road Planning and Road Design) was created in 1992 during the refoundation of the Department of Transportation Sciences "Friedrich List" at the Technische Universität Dresden. The field of work includes teaching and research on planning and design of roads. Since 2003 students are also involved in practical projects.
STEMMER IMAGING has been one of the leaders in the machine vision market since 1987. It is one of Europe's largest technology providers in this field. In 1997 STEMMER IMAGING presented Common Vision Blox (CVB), a powerful programming library for fast and reliable development and implementation of vision solutions, which has been deployed successfully throughout the world in more than 40,000 imaging applications in various industries.