Plastic covered vineyard extraction from airborne sensor data with an object-oriented approach

In recent years, the wide-spreading of vineyard cultivation in the Apulia Region (Italy) has showed negative consequences on the hydrogeological balance of soils as well as on the visual quality of rural landscape which has been significantly altered by the heavy diffusion of artificial plastic coverings. In order to monitor and manage this phenomenon, a detailed site mapping has become essential. With the increase of spatial resolution, pixel based approaches no longer capture the characteristics of classification targets. Consequently, classification accuracy is poor. Object-based image classification techniques overcome this issue by first segmenting the image into meaningful multipixel objects of various sizes and then assigning segments to classes using fuzzy methods and hierarchical decision keys. In this study an object-based classification procedure from Very High Spatial Resolution (VHSR) true color aerial data was developed on a test area located between the Apulian municipalities of Ginosa and Palagiano in order to support the update of the existing land use database aimed at plastic covered vineyard monitoring.