在利用遙測資訊進行水稻田分類問題當中,光譜反應值接近的相異類別往往容易產生混淆,例如水稻田、草地與林地等。也就是說若以光譜資訊進行影像判釋時,將很難直接獲得較高的分類精準度。為解決此一問題,本研究嘗試於影像判釋時加入紋理資訊,擴大植生類別間的空間特徵差異性,進而提升高解析度衛星影像判釋時的精準度。 本研究使用地理統計學(Geostatistic Theory)中的半變異元理論(Semivariogram Theory)作為紋理資訊之萃取模式,模式的選定則是使用方向半變異元(Direct-Semivariogram)與交叉半變異元(Cross-Semivariogram)模式,以表達出植生類別間不同性質之紋理特徵資訊。另外,研究中也深入的討論了視窗大小、計算方向等影響分類成果之因子特性,以得到適合輔助水稻田判釋之紋理資訊。研究成果顯示,利用紋理資訊可有效的提升高解析度衛星影像判釋水稻田的精確度,而分類成果也將有助於相關單位制訂農業政策時之重要參考依據。
The satellite image classification is one of the important application of remote sensing data. However, the vegetations (such as tree, grass and rice paddy) have very similar spectrum response in satellite images. This spectrum effect will caused to difficult classify each categories by using only mulitspectral data. To resolve this problem, this study used to the semivariogram theory (texture information) to extracted each vegetations feature of the QuickBird image. This research used the Direct-Semivariogram and Cross-Semivariogram to calculate image texture. Further, variogram analysis was performed on Quickbird data to determine the nature of spatial dependence with spectral reflectance for the selected vegetations of land cover systems which is estimated by the mean of the ranges in a series of variograms. On the other hand, this study also discussion the texture factor by the variogram analysis (such as window size, choose direction and band combination). Finally, the results showed semivariogram texture information which can effectively improve the classification accuracy in high resolution satellite image.