The Research of the Non-uniformity in Land Use Map Patches and Image Segments

Using multi-resource datasets to analyze remote sensing images is one of the focuses in the field of remote sensing image analysis. However, this technological theory is usually difficult to achieve due to the different structure and data standard of multi-resource datasets. This paper proposes a method which tries to analyze and solve the problem of the non uniformity in land use map patches and image segments. Image segments are obtained directly by matching land use map and images then sub segments are gotten by reclassifying image segments using supervised classification. Relations between land use map classes and sub segments are found by decision tree and then non uniformity of them could be resolved. The method used in this paper has to be considered prior to its operational use in analyzing land use map and Quickbird images in the region of Wuhan City in China.