The Integration of Hyperspectral and Radar Data to Enhance Urban Surface Features

Abstract

都市地物的種類多樣，透過單一感測器的資料往往無法有效判別地物特徵，本研究利用「明暗-色相-飽和度轉換法」（Intensity-Hue-Saturation transformation）來融合空載超光譜（AVIRIS）和雷達（TOPSAR）影像，以整合超光譜影像所提供的高光譜解析力和雷達影像所提供的地表組織與粗糙度等資訊於單一的影像中，並將此整合的影像疊合TOPSAR所提供的數值地形模型（DEM）資料，產生三維立體影像。對臨近山坡地的都市地區，此方法所產生的三維立體影像，由於同時具有地物和地形特徵，未來可被應用於模擬泥流或土石流災害。

Due to the complexity of urban surface materials, it is inefficient to discriminate urban land covers by a single sensor. In this project, the Intensity-Hue-Saturation (IHS) transformation is used to integrate the high spectral resolution, provided by the hyperspectral data (AVIRIS), and the surface texture information, derived from radar data (TOPSAR), into a single image. The spectral ambiguity of several ambiguity of several urban land covers is resolved in the merged image with higher spectral and spatial urban land covers is resolved in the merged image with higher spectral and spatial resolution. This image is also superimposed on the DEM data derived from TOPSAR data to create a 3-D perspective view. For the urban areas where are at risk of geological hazards (e.g., mudflows and debris flows), the 3-D perspective view of the fused image can be applied to simulate the movement of mudflows and predict the areas image can be applied to simulate the movement of mudflows and predict the areas and buildings at risk.