Boundary Extraction Method of Tidal Flat Region Based on Multi-spectral Image
The current tidal flat region boundary extraction method mostly adopts feature selection algo-rithm,which is vulnerable to non-boundary points,and can not effectively distinguish the region bounda-ry and shore features,resulting in low extraction accuracy.Therefore,a boundary extraction method for tidal flats based on multi-spectral images is proposed,which uses a multi-spectral imager to collect tidal flat images and obtains point cloud data in the tidal flat area based on the band distribution of the images.The point cloud data is gridded to identify the boundary point.The filtering principle is used to remove non boundary point,and the tidal flat regional boundary is extracted by calculating the elevation range of the regional grid.The experimental results show that the proposed method has an extraction accuracy of o-ver 95%for different tidal flat area boundaries,and the extraction time is within 5 seconds,which effec-tively realizes the boudaries extraction of the tidal flat area and improves the accuracy.
NETL
NSTL
万方数据
