Research of comparison of intelligent extraction methods based on high spatial resolution image

With the application of Geographic entities for the expression of remote sensing information model, Analysis of remote sensing image was improved from pixels to the object layers. Combined with rules and reasoning mechanism, identification of geographical entities had been improved with intelligence from the image objects using more multiple factors. The traditional pixel-based methods that could not been effectively carried out information extraction on the high spatial resolution image was broken. A result of comparison was Provided about information extraction of remote sensing image in a mining area of Guangxi. The classification accuracy of two methods based on pixels and objects was evaluated. The difficulty of integration had been reduced between remote sensing information and GIS.