MMS SPATIAL INFORMATION AUTOMATION EXTRACTION METHOD USING DEEP LEARNING TECHNIQUE

摘要

The present invention relates to a mobile mapping system (MMS) spatial information automation extraction method and, more specifically, to an MMS spatial information automation extraction method applied with a deep learning technique capable of effectively removing a shielding area. The method according to the present invention comprises: a representative object patterning step of patterning a plurality of representative objects based on MMS data; an object extracting step of extracting an object from the acquired MMS data; a first shielding object selecting step of selecting a first shielding object having a shielding area among the objects extracted in the object extracting step; a shielding area extracting step of extracting a shielding area from the first shielding object; a shielding rate calculating step of calculating a shielding rate of the first shielding object; a second shielding object selecting step of selecting a second shielding object having a shielding rate within a reference value range by comparing the shielding rate with a preset reference value; a similar representative object selecting step of selecting an object similar to the second shielding object from the plurality of patterned representative objects; a shielding space processing step of processing shielding spatial information of the second shielding object by comparing the selected similar representative object with the second shielding object; and a shielding spatial information automation processing list storing step of storing the shielding spatial information-processed second shielding object in a shielding spatial information automation processing list.;COPYRIGHT KIPO 2020