Extracting Farmland Features from Lidar-Derived DEM for Improving Flood Plain Delineation

摘要: Flood plains, which are commonly distributed in flat river or lake basins, often contain large tracts of farmland. Therefore, flood plains require precise and detailed information on the role played by farmland routing simulations, risk evaluation, loss evaluation. In farmland, cultivated land parcels not directly adjacent. The intervening non-cultivable land, might include trails ditches, can cover areas. Currently, area between is usually measured artificial visual interpretation fieldwork. This study focused extraction uncultivable trails, field within basis a Light Detection Ranging-derived (LiDAR-derived) high-resolution gridded Digital Elevation Model (DEM). proposed approach was applied to generate polygons individual storage detention area. DEM first smoothed then subtracted. To remove small spots smooth boundaries parcels, inner outer buffers were created generalize extracted polygons. Experiments proved that this applicable plain demonstrated chosen parameters appropriate. more efficient than traditional surveying methods. For parcel extraction, accuracy achieved 93.42%, using official statistics for comparison, Cohen’s kappa coefficient 0.90, an aerial image comparison. coefficients 0.87 0.77 trail ditch respectively.