Lot-edge extraction from cadastral map using object-oriented classification
Tan Qulin · Journal of Beijing Jiaotong University · 2013
A scheme based on object-oriented classification analysis was developed to extract lot-edge from scanned cadastral map in the paper.Firstly,segmentation was performed to form image objects or regions in the digital raster map.Secondly,four knowledge rules incorporated into a fussy logic classifier were constructed for classifying lot and non-lot image objects.Finally,lot polygons were extracted by vectorization and generalization.