Identifying ephemeral gullies from high-resolution images and DEMs using flow-directional detection

摘要: Ephemeral gullies, which are widely developed worldwide and threaten farmlands, have aroused a growing concern. Identifying mapping gullies generally considered prerequisites of gully erosion assessment. However, ephemeral remains challenge. In this study, we proposed flow-directional detection for identifying from high-resolution images digital elevation models (DEMs). exhibit clear linear features in images. An edge operator was initially used to identify Then, according mechanism, the designed. Edge obtained flow directions DEMs were implement that detects along direction. Results ten study areas Loess Plateau China showed ranges precision, recall, F-measure 60.66%–90.47%, 65.74%–94.98%, 63.10%–91.93%, respectively. The method is flexible can be with various DEMs. analysis effect DEM resolution accuracy only demonstrates minor on results. Conversely, influences result more important than resolution. worse vertical DEM, lower performance will be. This work beneficial research related monitoring assessing soil loss.