Echo Detection and Land Cover Classification of Airborne Waveform LiDAR Data

空載光達波形資料之響應偵測與土地覆蓋分類

王正楷(Cheng-Kai Wang) ； 曾義星(Yi-Hsing Tseng)

《航測及遙測學刊》 21卷2期 (2016/10) Pp. 75-94

https://doi.org/10.6574/JPRS.2016.21(2).2

摘要

地形製圖的製作，在眾多的應用中，已越來越重要，例如災害防救、森林資源管理、海岸的侵蝕等等皆需要地形圖來輔助決策。為了製作一張地形圖，通常需要地面的高程起伏資訊以及地表面上的地物類別來共同組成。本研究的目的乃針對空載光達波形系統所獲取的資料，探究其製作地形圖的潛力，並與傳統離散光達系統相比較。本研究首先從波形資料中偵測在傳統離散光達系統往往忽略掉的點雲，藉此獲得較好的場景幾何描述較佳的點雲，透過這較佳的點雲資料，有助於提升自動化生產的數值高程模型之精度。另外，我們對於波形資料所隱含的特徵，透過一系列的訊號處理方式，將其轉換為有用的波形特徵，並用以進行土地覆蓋分類。

關鍵字

空載光達 ； 波形 ； 數值高程模型 ； 土地覆蓋

並列摘要

Mapping of geospatial earth surfaces has become increasingly important and been needed in many applications, such as disaster prevention, forest management, and monitoring of coastal erosion and sediment transport. Modeling the topography of bare earth and identifying land covers are essential for mapping the earth surfaces. This paper investigates modern small-footprint full-waveform airborne LiDAR systems for the mapping of geospatial earth surfaces in order to provide solutions to difficulties that a conventional discrete-return airborne LiDAR system often encounters. The research begins with the detection of points which are often missed by a discrete-return LiDAR system. The generation of a digital elevation model is then benefited from the enhanced geometry of landscapes. Another significant advantage of full-waveform data is that additional information characterizing different surface types can be extracted. Thus this research continues to identify the earth land covers using information extracted from waveform data.

並列關鍵字

airborne LiDAR ； waveform ； digital elevation model ； land cover