Typical Objects Extraction From LiDAR Data In Coastal Zone

The investigation of land use in coastal zone plays an important role in the rational exploitation and application of coastal zone.It is more significant to investigate survey the distribution of water regions,buildings and roads,which reflects the economic and social development of the coastal zone.Airborne Light Detection and Ranging System(Lidar)can obtain the three-dimensional terrain information timely and accurately,which displays an extremely advantage in land use classification.While there were few researchers focus on the studies in the extraction of water regions in costal zone.Furthermore,existing methods for extracting water regions can not be applied to coastal zone directly on account of the complex distribution of waters.Therefore it is of great significance to extract water regions,buildings and roads in coastal zone with Li DAR data.The characteristics water regions,buildings and roads in Li DAR data are comprehensively analyzed in the paper and a new extraction process was proposed:(1)The water regions were obtained by integrating the elevation,intensity and point-density information derived from the highly accurate 3D mass points of airborne Li DAR.Firstly,the initial water points were extracted through using the feature of low density;These water points were then filtered by the elevation and intensity threshold.Secondly,A triangulation network surface model was established based on the water points to describe the elevation trend of the water surface.Lastly,the final water points were extracted by the relative elevation between the Points Cloud and Triangulation.(2)The building points were extracted from the remaining points after water extraction.Points cloud filtering was executed to classify ground and non-ground data.Then,those points which are higher than ground over the threshold were extracted.Building points were detected by the planarity of point groups and be optimized by spatial relationships.(3)The road points were extracted from the ground points.Road point samples were selected from ground points.Then the road intensity threshold was set from the statistical distribution of those sample points.The points in the scope of the road intensity threshold were extracted and the noise and mudflat was removed by density feature.Lastly,road points were extracted by the connectivity feature of roads.