Roads have multiple effects on wildlife, from animal mortality, habitat and population fragmentation, to modification of animal reproductive behavior. Amphibians in particular, due to their activity patterns, population structure, and preferred habitats, are strongly affected by traffic intensity and road density. On the other hand, road-kills studies and conservation measures have been extensively applied on highways, although amphibians die massively on country roads, where conservation measures are not applied. Many countries (e.g. Portugal) have not any national program for monitoring road-kills, a common practice in other European countries (e.g. UK; The Netherlands). This is necessary to identify hotspots of road-kills in order to implement conservation measures correctly. However, monitoring road-kills is expensive and time consuming, and depend mainly on volunteers. Therefore, cheap, easy to implement, and automatic methods for detecting road-kills over larger areas (broad monitoring) and along time (continuous monitoring) are necessary. We present here the preliminary results from a research project which aims to build a cheap and efficient system for detecting amphibians roadkills using computer-vision techniques from robotics. We propose two different solutions: 1) a Mobile Mapping System to detect automatically amphibians’ road-kills in roads, and 2) a Fixed Detection System to monitor automatically road-kills in a particular road place during a long time. The first methodology will detect and locate road-kills through the automatic classification of road surface images taken from a car with a digital camera, linked to a GPS. Road kill casualties will be detected automatically in the image through a classification algorithm developed specifically for this purpose. The second methodology will detect amphibians crossing a particular road point, and determine if they survive or not. Both Fixed and Mobile system will use similar programs. The algorithm is trained with existing data. For now, we can only present some results about the Mobile Mapping System. We are performing different tests with different cameras, namely a lineal camera, used in different industrial solutions of control quality, and an outdoor Go-pro camera, very famous on different sports like biking. Our results prove that we can detect different road-killed and live animals to an acceptable car speed and at a high spatial resolution. Both Mapping Systems will provide the capacity to detect automatically the casualties of road-kills. With these data, it will be possible to analyze the distribution of road-kills and hotspots, to identify the main migration routes, to count the total number of amphibians crossing a road, to determine how many of that individuals are effectively road-killed, and to define where conservation measures should be implemented. All these objectives will be achieved more easily at with a lower cost in funds, time, and personal resources.
