A classification model of grazing areas in southern Italy using Remote Sensing data.

Detection of vegetation typologies is of particular interest in epidemiological studies on animal diseases. This paper is focused on the automatic detection of grazing areas in southern Italy browsed by equines, oxen and sheep, mostly sedentary, with many in fenced areas for overnight stays. Results revealed Satellite Remote Sensing was an indispensable tool in area surveys for vegetation cover characterization. Our classification model shows an accuracy level of 90.21% and a precision of 92.69%. Results suggest similar application protocols can be used in areas with different vegetation cover to characterizing potential infection risk areas for geohelminths and other diseases.