透過您的圖書館登入
IP:3.144.42.196
  • 期刊

福衛二號影像之大氣輻射校正-輻射控制區的應用

Atmospheric Correction of Formosat-II Imagery - Application of Radiometric Control Areas

摘要


衛星遙測於地表資源監測上是相當具有潛力且快速發展中的工具之一,尤其福衛二號影像具有八公尺空間解析度,被廣泛應用於農林業與水資源監測等各領域。大氣層對太陽入輻射產生散射作用而被衛星所接收之輻射量(或稱為路徑輻射量),與地表物體反射特性無關,因此欲使用衛星影像準確推估地表特性,必須適當地移除路徑輻射量,對於後續應用於農業與水資源領域而言相當重要。本文提出基於輻射控制區中實際量測之反射率估算路徑輻射量的流程,並與傳統上常用方法比較,分別是暗體辨識法、多波段迴歸法與共變矩陣法等。研究中顯示柏油鋪面適合作為輻射控制區。四種方法的結果顯示,暗體辨識法與共變矩陣法容易高估路徑輻射量,而多波段迴歸法雖然得到與本文所提方法相似的結果,但是由於偏低的判定係數使得迴歸模式的可靠度較低。綜合三個間接評估指標的結果,本文所提出之基於輻射控制區的方法均優於其他傳統方法。

並列摘要


Remote sensing has been extensively applied to Earth resource monitoring and management. However, radiances received at satellite sensors include not only radiances reflected from the target on land surface but also radiances scattered by particles in the atmosphere, i.e. the path radiance. Path radiance is independent of the target characteristics and its presence induces noise to satellite images. In this paper, an algorithm of path radiance estimation based on measurements of surface reflectance at radiometric control areas (RCA) is proposed. Path radiance estimates of the proposed RCA-based method were compared against estimates of other methods including the dark object subtraction method, multi-band regression technique, and the covariance matrix method. The RCA-based method is superior to other methods based on three assessment criteria. In this research, paved surface which could be easily identified in most images was found to be a good choice for radiometric control areas. The dark object subtraction method and the covariance matrix method tend to overestimate path radiance. Although in our study the multi-band regression technique seems to perform as well as the RCA-based method, its estimates may be less reliable.

延伸閱讀