Experiment about image emporium buildings intelligent recognition using spatial semantic model

Automatically extraction model, which is based on shape, spectrum, texture and other characteristics information, achieves a better effect in the information extraction of the buildings, but it is difficult to identify the building further. Therefore, this paper proposes we should consider the spatial association features and semantic features characteristic of the objectives when identifying the types of the buildings. This paper takes the emporium construction as the experiment objectives. In the base of analyzing the integrated features of the emporium construction, building emporium and construction spatial semantic model, automatically extracting the emporium construction information from the high-resolution remote sensing image and then testing and analyzing the recognition results. The result shows that this method is an effective way to automatically identify the building classification from the remote sensing image. ©2009 IEEE.