Mobile mapping system (MMS2) for detecting roadkills.

Roads affect negatively wildlife, from direct mortality to habitat fragmentation. Mortality caused by collision with vehicles on roads is a major threat to many species. Monitoring animal road-kills is essential to stablish correct road mitigation measures. Many countries have national monitoring systems for identifying mortality hotspots. We present here an improved version of the mobile mapping system (MMS2) for detecting Roadkills not only for amphibians but small birds as well. It is composed by two stereo multi-spectral and high definition camera (ZED), a high-power processing laptop, a GPS device connected to the laptop, and a small support device attachable to the back of any vehicle. The system is controlled by several applications that manage all the video recording steps as well as the GPS acquisition, merging everything in a single final file, ready to be examine by an algorithm at posterior. We used the state-of-the-art machine learning computer vision algorithm (CNN: Convolutional Neural Network) to automatically detect animals on roads. This self-learning algorithm needs a large number of images with alive animals, road-killed animals and any objects likely to be found on roads (e.g. garbage thrown away by drivers) in order to be trained. The greater the image database, the greater the detection efficiency. This improved version of the mobile mapping system presents very good results. The algorithm has a good effectiveness in detecting small birds and amphibians.

Data and Resources

Cite as

Ribeiro H. Sillero N. y Guedes D. Mobile mapping system (MMS2) for detecting roadkills. Infrastructure & Ecology Network Europe, 2021.

Clipboard Icon
Retrieved: 18 Jan 2025 14:32:02

Metadata
Download
DCAT-AP/NTI-RISP

Basic information
Resource type Text
Date of creation 2024-12-02
Date of last revision 2025-01-18
Show changelog
Metadata identifier f8fb6d24-5bc3-54b5-9cd8-0a386c43ea0f
Metadata language Spanish
Themes (NTI-RISP)
High-value dataset category
ISO 19115 topic category biota
Keyword URIs
Bibliographic information
Name of the dataset creator Ribeiro, H., Sillero, N. y Guedes, D.
Name of the dataset editor Infrastructure & Ecology Network Europe
Other identifier ISBN: 978-972-778-182-9
Identifier of the dataset creator
Email of the dataset creator
Website of the dataset creator
Provenance
Lineage statement
Metadata Standard
Version notes
Version