數據整合方法應用於SPOT衛星影像之分析研究

Abstract

隨同電腦影像處理、分析及地理資訊系統之技術發展，多樣性資料整合的 應用，已日趨重要。各種不同性質、種類、來源及時間的資料可提供不同 的訊息。如何整合不同類型之資料，使之相輔相成，乃目前研究的一大課 題。其中法國的資源衛星SPOT為一個很好的資訊來源。本文就以SPOT的衛 星影像資料為試驗對像，進行不同整合方法的比較研究。本研究針對：一 、色彩模型轉換法，二、主軸分析法，三、高頻通過濾波法，四、輻射量 度法等四種方法來進行探討。整合後成果的評估則以統計法中的相關係數 法、均方根誤差法、影像像元灰值差異法等方法來評估影像的差異量，以 熵值的計算和灰值分佈圖的比較來評估整合後資訊量的增加情形，並以視 覺的方法直接判釋影像內容。 The integration of differebt data sorce inGIS has received increasing emphasis as a result of new development in computer and GIS techgnology. Integration of various data sets to fully utilize complementary information hsa become an important component. In recent years, there has been a large increase in the amount of image data that are available to users for various purposes. The launch of the French SPOT satellite system has given the capability to a range of land use and land cover analyses. This thesis compares the results of diffenent methods used to integrate the information contents of the SPOT Panchromatic and Multispectral image data. Four integrating methods, namely, clour space transformation, principal component analysis, high pass filter and radiometric method are ivvestigated. To evaluate the result of different merging methods, six measures are used. They are visual inspection, correlation, root mean square error, difference of two images, entropy value and the histogram comparison.