i-MeaS

Worldwide rock falls are one of the major types of natural hazards killing or injuring a large number of individuals and creating very high costs every year. In the United States of America such events are annually causing estimated damages exceeding US $ 2 billion. Figures for Europe, Asia, particular China, Africa and South America may easily exceed those for the US. Besides direct costs rock falls are also reason for even higher indirect costs like interruption of important infrastructure facilities or losses for the tourist industry etc. In future it is very likely that the damages caused by rock falls will even increase as the hilly areas, where the majority of the rock falls occur, are used by a growing number of tourists and intersected by increasingly powerful transnational networks. In addition many global climate change scenarios predict an increase in the probability for heavy rain, which is a primary trigger for rock falls and landslides. This implies that there is urgent need for highly productive and reliable tools for rock fall monitoring at an operational level. i-MeaS aims at an efficient, highly automated intelligent image-based measurement and analysis system which is able to monitor rock falls by means of non-signalised points. Basic idea is the use of image analysis techniques for automated point detection. In connection with such algorithms a huge number of constraints have to be clarified and solved (e.g. the influence of different light conditions on the resulting measurements) - we plan to combine image analysis techniques, with techniques from artificial intelligence (e.g. learning algorithms, knowledge-based systems, genetic algorithms, etc.) to research and assemble a flexible measurement system. Additionally, the system should be integrated into an alerting system. The investigation and installation of such a combined monitoring systems aims for a raise of security and a restriction of human, economical and environmental damage. The innovation of i-MeaS is the close interaction between feature extraction, image-based sensors and a deformation analysis/alerting system. In traditional measurement systems sensors are used as (passive) devices only. Besides the input from image analysis there are complementary input data making it possible to build up activities of the sensors in feedback circuits. Totally new is the development and implementation of such a measurement system for the task of rock fall monitoring. The utilization of such a multi-sensor system has several advantages in comparison with conventional systems, like laser scanner or tacheometers. The project at hand is planned as an international and interdisciplinary work and is designed as a "FWF Translational-Research-Programm". The project can refer to several research projects executed by the project proposers.