Towards the automatic olive trees extraction from aerial and satellite imagery

In this paper an approach towards the automatic olive tree extraction from satellite imagery is presented. Counting olive trees in specific agriculture areas is crucial for the European Union olive production subsidies. Still counting and verification of each producer’s subsidy are based on standard photointerpretation methods or semi automatic techniques. In this paper an image processing scheme towards automatic olive tree extraction was developed and applied to satellite imagery. The processing scheme is composed of two steps. In the first step, enhancement and smoothing was performed using nonlinear diffusion. In the second step, olive tree extraction was accomplished after extracting the local spatial maxima of the Laplacian. Results appear promising and counting errors mostly occur in cases where olive trees can not even be detected well by standard photointerpretation methods