From pixel to vine parcel: A complete methodology for vineyard delineation and characterization using remote-sensing data

8 The increasing availability of Very High Spatial Resolution images enables accurate 9 digital maps production as an aid for management in the agricultural domain. In 10 this study we develop a comprehensive and automatic tool for vineyard detection, 11 delineation and characterization using aerial images and without any parcel plan 12 availability. In France, vineyard training methods in rows or grids generate periodic 13 patterns which make frequency analysis a suitable approach. The proposed method 14 computes a Fast Fourier Transform on an aerial image, providing the delineation 15 of vineyards and the accurate evaluation of row orientation and interrow width. 16 These characteristics are then used to extract individual vine rows, with the aim of 17 detecting missing vine plants and characterizing cultural practices. Using the red 18 channel of an aerial image, 90% of the parcels have been detected (56.2% with cor19 rect boundaries); 92% have been well classified according to their rate of missing 20 vine plants and 81% according to their cultural practice (weed control method). The 21 automatic process developed can be easily integrated into the final user’s Geograph22 ical Information System and produces useful information for vineyard management. 23