Automatic spatial data acquisition from scanned mine maps based on mathematical morphology

摘要

A series of algorithm models for extracting cartographic data from scanned mine maps are probed and practiced based on set theory and amthematical morphology. These models include dynamic adaptive binarizating with multi threshold, noise removal, extraction of multi layers of cartographic elements, shape preserving image skeletonization and nodes searching. The proposed algorithms are tested on a portion of a scanned surface-underground mine map. After different line elements are separated into multi layers and skeletonized, these line elements are ready for converting to vector format and inputting to MGIS. All the test results show that these algorithms work efficiently.