Do Crash Barriers and Fences Have an Impact on Wildlife-Vehicle Collisions? - An Artificial Intelligence and GIS-Based Analysis

Wildlife–vehicle collisions (WVCs) cause significant road mortality of wildlife and have led to the installation of protective measures along streets. Until now, it has been difficult to determine the impact of roadside infrastructure that might act as a barrier for animals. The main deficits are the lack of geodata for roadside infrastructure and georeferenced accidents recorded for a larger area. We analyzed 113 km of road network of the district Freyung-Grafenau, Germany, and 1571 WVCs, examining correlations between the appearance of WVCs, the presence or absence of roadside infrastructure, particularly crash barriers and fences, and the relevance of the blocking effect for individual species. To receive infrastructure data on a larger scale, we analyzed 5596 road inspection images with a neural network for barrier recognition and a GIS for a complete spatial inventory. This data was combined with the data of WVCs in GIS to evaluate the infrastructure’s impact on accidents. The results show that crash barriers have an effect on WVCs, as collisions are lower on roads with crash barriers. In particular, smaller animals have a lower collision share. The risk reduction at fenced sections could not be proven as fenced sections are only available at 3% of the analyzed roads. Thus, especially the fence dataset must be validated by a larger sample number. However, these preliminary results indicate that the combination of artificial intelligence and GIS may be used to analyze and better allocate protective barriers or to apply it in alternative measures, such as dynamic WVC risk-warning.

[1]  Luis Moreno,et al.  Road traffic sign detection and classification , 1997, IEEE Trans. Ind. Electron..

[2]  Clayton K. Nielsen,et al.  Landscape influences on deer-vehicle accident areas in an Urban environment , 2003 .

[3]  A. Clevenger,et al.  Potential Impacts of Highway Median Barriers on Wildlife: State of the Practice and Gap Analysis , 2013, Environmental Management.

[4]  M. Lizana,et al.  Factors influencing wolf Canis lupus roadkills in Northwest Spain , 2011, European Journal of Wildlife Research.

[5]  Johannes Stallkamp,et al.  Detection of traffic signs in real-world images: The German traffic sign detection benchmark , 2013, The 2013 International Joint Conference on Neural Networks (IJCNN).

[6]  R. Warren,et al.  Movements and home ranges of white-tailed deer in response to roadside fences† , 2011 .

[7]  D. L. Kellmeyer,et al.  Detection of highway warning signs in natural video images using color image processing and neural networks , 1994, Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94).

[8]  Jukka Matthias Krisp,et al.  Segmentation of lines based on point densities--an optimisation of wildlife warning sign placement in southern Finland. , 2007, Accident; analysis and prevention.

[9]  Christoph Mertz,et al.  Vision for road inspection , 2014, IEEE Winter Conference on Applications of Computer Vision.

[10]  Giorgos Mountrakis,et al.  Spatial wildlife-vehicle collision models: a review of current work and its application to transportation mitigation projects. , 2011, Journal of environmental management.

[11]  Majid Mirmehdi,et al.  Real-Time Detection and Recognition of Road Traffic Signs , 2012, IEEE Transactions on Intelligent Transportation Systems.

[12]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[13]  Maria Castro,et al.  Using smartphones as a very low-cost tool for road inventories , 2014 .

[14]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[15]  Z. Keken,et al.  Landscape structural changes between 1950 and 2012 and their role in wildlife–vehicle collisions in the Czech Republic , 2016 .

[16]  Ayman Alfalou,et al.  Road sign identification and geolocation using JTC and VIAPIX module , 2018, Defense + Security.

[17]  Trisalyn A. Nelson,et al.  A time geographic approach for delineating areas of sustained wildlife use , 2015, Ann. GIS.

[18]  Jianwei Gong,et al.  Mobile Terrestrial Laser Scanning for Highway Inventory Data Collection , 2012 .

[19]  SPATIOTEMPORAL ANALYSIS FOR WILDLIFE-VEHICLE-COLLISIONS BASED ON ACCIDENT STATISTICS OF THE COUNTY STRAUBING-BOGEN IN LOWER BAVARIA , 2016 .

[20]  Francisco Suárez,et al.  Can we mitigate animal–vehicle accidents using predictive models? , 2004 .

[21]  Jesmin F. Khan,et al.  Image Segmentation and Shape Analysis for Road-Sign Detection , 2011, IEEE Transactions on Intelligent Transportation Systems.

[22]  David A Noyce,et al.  Development of Data Collection and Integration Framework for Road Inventory Data , 2010 .

[23]  Manuel Lagunas,et al.  Transfer Learning for Illustration Classification , 2018, CEIG.

[24]  A. Seiler.  Predicting locations of moose–vehicle collisions in Sweden , 2005 .

[25]  Sei-Wang Chen,et al.  Road-sign detection and tracking , 2003, IEEE Trans. Veh. Technol..

[26]  Angela Kociolek,et al.  Highway median Impacts on Wildlife Movement and Mortality , 2006 .