以地面光達資料重建大葉桃花心木三維模型

Using Ground-based LiDAR Data to Reconstruct the Three Dimensional Model of Swietenia macrophylla King

蔡馨儀(Hsin-Yi Tsai)

指導教授 ： 陳朝圳

國立屏東科技大學/農學院/森林系所/碩士(2012年)

https://doi.org/10.6346/NPUST.2012.00183

摘要

傳統森林資源調查，藉由破壞性量測進行林木測計資料之收集，會花費大量人力與成本以達到森林調查之目的。在科技快速進步下，目前結合遙測技術於森林調查，已成為趨勢，但遙測資料容易遭受氣候與地形遮蔽影響，使得遙測資訊提供受到限制。光達系統是一種高科技、高精度、低成本、快速及非破壞性量測技術，在國外研究證實光達系統能精確量測林木性態值包括直徑、樹高材積和生物量等，並可利用其三維資訊進行林木及林分狀態之模擬。本研究以大葉桃花心木人工林為研究材料，以地面光達資料探討林木性態值量測之可行性與林木三維模型建構之應用。本研究於2010年與2011年設置15 m x 55 m長方形之大葉桃花心木人工林樣區，使用Mensi GS 100與FARO Photon 120進行林木之地面光達掃描，以Rapidform XOR3萃取單木林木性態值資料，並以野外實測調查資料，探討地面光達於林木測計的準確性，並以光達林木性態值資料，推估單木及林分連年生長量，最後以光達三維空間資訊特性，使用逆向工程技術重建大葉桃花心木三維模型，應用於研究樣區景觀模擬。結果顯示以地面光達掃描之表面點雲三維資訊於林木性態值的準確度高，故以地面光達於對於坡度平緩林地與同齡林分推估林分性態值可行。地面光達具有高準確度與非破壞性量測的優勢，在林木性態值生長及連年生長量推估，可明顯看出林木性態值一年間變化情形，但在於樹冠幅方面，推估林木間重疊不易看出生長變化。連年生長量推算結果2010年林分總蓄積量為51.52 m3/ ha，2011年為98.60 m3/ ha，其連年生長量為47.08 m3/ ha。地面光達除了能精準掃描林木表面，獲取大量三維點雲資料外，能藉由逆向工程技術，重建大葉桃花心木實體三維模型，展示森林景觀視覺模擬情形。

關鍵字

地面光達 ； 三維雷射掃描 ； 林木性態值 ； 林木三維模型

並列摘要

Traditional forest inventory that collected the tree measurement data by destroyed measurements would cost huge manpower and cost. Because of the progression of technology, applying remote sensing techniques to forest inventory becomes current trends. However, remote sensing information is limited due to the incorrect remote sensing data which is affected by climate and terrain shadow. LiDAR system is a high technology, high precision, low cost, and non-destroyed measurement techniques. LiDAR system could not only precisely measures trees attributes, such as DBH, height, volume and biomass, but also simulates trees and stand status by using three dimensional information in foreign studies. This study used Swietenia macrophylla King artificial forests as a material to explore the availability of trees attribute measurements and the application of trees three dimensional modeling by using ground-based LiDAR data. Firstly, we set up rectangular Swietenia macrophylla King artificial forest sample area which is 15m X 55m in 2010 and 2011. Then we did ground LiDAR scanning via Mensi GS 100 and
FARO Photon 120, and retrieved trees attribute data via Rapidform XOR3. In addition, field data survey was used to investigate the accuracy of applying ground-based LiDAR to forest measurement. We also used LiDAR trees attribute data to evaluate individual tree and current annual increment. Lastly, reversed engineering was be utilized to reconstruct the tree dimensional model of Swietenia macrophylla King that would be apply to landscape modeling in study sample areas. The results showed the high accuracy of using surface cloud point information of ground-based LiDAR scanning to evaluate trees attribute. Hence, it is feasible to evaluate stand attribute in flat slope forest land and even aged forest via ground-based LiDAR. Ground-based LiDAR has the advantages of high accuracy and non-destroyed measurement, so it can be used to evaluate trees attribute and current annual increment obviously. Nonetheless, in terms of tree crown rate, it is not easy to evaluate variety. Current annual increment of stand was 47.08 m3/ ha (Per hectare stocking was 51.52 m3/ ha in 2010, 98.60 m3/ ha in 2011). In conclusion, ground-based LiDAR not only scan trees precisely and get lots of three dimensional cloud point data, but also reconstruct the three dimensional model of Swietenia macrophylla King and display forest landscape visual modeling via reversed engineering techniques.

並列關鍵字

Ground-based LiDAR ； 3D Laser Scanning ； Tree characteristics ； 3D tree model

參考文獻

魏浚紘 (2008) 應用空載光達推估阿里山地區柳杉人工林林分材積。屏東科技大學森林系所碩士論文，106頁。

汪大雄、謝漢欽 (2010) 森林景觀視覺化在森林經營上之應用。林業研究專訊 17(1): 49-52。

蕭淳伊、曾義星 (2009) 應用遙測影像與空載光達資料推估森林分佈面積及樹冠體積。航測及遙測學刊 14(1): 51-64。

陳永寬、詹進發、葉堃生、鄭祈全、施瑩瑄 (2005) 應用地面雷射掃描儀測算樹高之研究。航測及遙測學刊 10(4): 327-336。

郭彥甫、林達德、王嘉銳 (2004) 應用影像處理技術輔助植物葉片之幾何模擬。農業機械學刊 13(2): 1-17。