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.

Datos y Recursos

Cite como

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

Clipboard Icon
Recuperado: 18 Jan 2025 20:38:23

Metadatos
Descargar
DCAT-AP/NTI-RISP

Información básica
Tipo de recurso Texto
Fecha de creación 02-12-2024
Fecha de última modificación 18-01-2025
Mostrar histórico de cambios
Identificador de los metadatos f8fb6d24-5bc3-54b5-9cd8-0a386c43ea0f
Idioma de los metadatos Español
Temáticas (NTI-RISP)
Categoría del conjunto de alto valor (HVD)
Categoría temática ISO 19115 biota
URI de palabras clave
Información bibliográfica
Nombre del autor Ribeiro, H., Sillero, N. y Guedes, D.
Nombre del editor Infrastructure & Ecology Network Europe
Identificador alternativo ISBN: 978-972-778-182-9
Identificador del autor
Email del autor
Web del autor
Procedencia
Declaración de linaje
Perfil de Metadatos
Notas sobre la versión
Versión