Campagnolo, Manuel Lameiras FigueiredoSchmedtmann, Jonas2015-03-242015-03-242014Schmedtmann. J. - Automatizing photo interpretation of satellite imagery in the context of the Common Agriculture Policy subsidy control. Lisboa: ISA, 2014, 64 p.http://hdl.handle.net/10400.5/8294Mestrado em Engenharia do Ambiente - Instituto Superior de AgronomiaComputer Assisted Photo-Interpretation (CAPI) uses remotely sensed imagery to control farmers’ subsidy applications in the context of the EU’s Common Agriculture Policy. A simple and reproducible method to automatize CAPI in an operational context with the overreaching goal to reduce control costs and completion time was developed in this study. Validated control data provided by the Portuguese Control and Paying Agency for Agriculture (IFAP) and a multispectral atmospherically corrected Landsat ETM+ time series were used to calibrate and test the method. Taking advantage of the nature of subsidy declarations, object-based land cover classification for the 12 most controlled classes was carried out in the region of Ribatejo. The main feature of the presented method is that it allows choosing a confidence level on the automatic classification of farmers’ parcels. While higher confidence levels reduce the risk of misclassifications, lower levels increase the number of automatic control decisions. A confidence level of 80% is a good compromise. This confidence level leads to over 55% of automatically taken control decisions with an overall accuracy of 84%. Furthermore, over 85% of all parcels classified as maize, rice, wheat or vineyard can be controlled by the method with the optimal confidence level.porCommon Agricultural Policysubsidy controlLandsatmultitemporal analysisoperational crop discriminationparcel-based classificationremote sensingmaster thesis201846772