A point bar is a sediment deposition that forms on river banks as a result of river sedimentation.Changes in point bars are regarded as an indirect response to the evolution of a river's morphology and its surrounding environment.Therefore,dynamic monitoring of point bars is significant for river-ecosystem research and shipping-lane management.It is difficult to distinguish point bars from some buildings due to the similarity of spectral characteristics.In this study,an innovative method to identify and extract point bars was proposed.This method was based on the multifarious differential characteristics such as spectrum,space,and phenology of the river and point bars in Landsat images.A binary change detection in multi-temporal water indices and an object-based image analysis were used in this method.The point bars in the middle and lower reaches of Gan river were identified and extracted by it.This method,on the one hand,enriches the methodology of monitoring point bars and provides a new perspective for the geographical study of point bars,and on the other hand,provides abundant data for the ecosystem construction,environmental management and shipping-lane planning in the Gan river basin.
point barOBIAchange detection difference analysismiddle and lower reaches of Gan riverLandsat
NETL
NSTL
万方数据
