Study of RS data classification based on rough sets and C4.5 algorithm
Ming Yu, Tinghua Ai · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 2009
The classification by extracting of remote sensing (RS) data is the primary information source for GIS in land resource application. Automatic and accurate mapping of region LUCC from high spatial resolution satellite image is still a challenge. The paper discussed remote sensing image data classification techniques based on C4.5 algorithm and rough sets and the combination of C4.5 algorithm and rough sets. On the basis of the theories and methods of spatial data mining, we improve the classification accuracy. Finally validates its effectiveness taking a test area as example. We took the outskirts of Fuzhou with complicated land use in Fujian Province as study area. The classification rules are discovered from the samples through decision tree C4.5 algorithm, Rough Sets and both with together, which integrates spectral, textural and the topography characters. And the classification test is performed based on these rules. The traditional maximum likelihood classification is also compared to check the classification accuracy. The results have shown that the accuracy of classification based on knowledge is markedly higher than the traditional maximum likelihood classification. Especially the method based on combine Rough Sets and decision tree C4.5 algorithm is the best.