應用全景圖像於量測森林樹冠之探討

Abstract

在森林生態系中，樹冠結構扮演著重要的角色，因為其具有多樣的功能性，如:影響初級生產力、土壤、碳交換、天氣系統、以及光輻射的能量等，因此，為了能夠更精確地計算、了解樹冠在生態系中是如何影響及其影響的量化數據，已有多數的研究方式及儀器測量，包含了LAI-2200、LiDAR、衛星影像、半球面相片等，除上述方式，目前已有高解析度全景影像應用在森林監測上，但整體在森林量測上的研究並不多，故本研究將建立高解析度全景影像於林下之拍攝詳細流程並探討全景影像應用於森林樹冠。試驗地主要為森林資源豐富、林相多樣的溪頭自然森林園區，樣點將分別設置針葉林、闊葉林、竹林及綠地公園四種林相中，並再挑選冠層鬱閉疏密程度不同的樣區，最後與技術成熟之半球面影像在同一樣點上拍攝，並使用影像分析軟體計算兩種影像之冠層參數與光立地係數，以單一樣品均值推論、簡單線性回歸與雙因子變異數方析法評估兩種影像應用於樹冠量測之差異。試驗結果顯示，全景與半球面影像兩種拍攝模式，分別透過SOLARCALC7.0與ArcGIS所獲取之數值，經上述統計計算結果顯示無顯著差異，且相關性高，表示高解析度全景影像能應用於森林樹冠之光輻射量測；而SOLARCALC7.0與ArcGIS兩種軟體，同時針對單一拍攝模式分析，其獲取之數據亦無明顯差異，且相關性高。建議進行森林影像拍攝，可採用全景影像，因其仍具有優勢，高解析度提供視覺判釋，如棲地破壞與回復與物候調查，未來更能應用於樹木定位與測量、林分取樣及材積推估等，而影像分析軟體則可直接使用SOLARCALC7.0進行分析。

In the forest ecological system, the canopies play an important role. Because it contains many funcations, like that affects net primary production (NPP), soil carbon exchange, weather system, and the energy of light radiation and so on. Therefore, for calculating the canopies of vegetation precisely and understanding the canopies how to affect the ecological system and what the quantitative data, there have been many study methods and instruments about measuring the canopies such as LAI-2200, LiDAR, satellite image and hemispherical images and so on. In addition to the above-mentioned methods, high resolution panoramic images now is applied to forest monitoring. But it is not much studies on forest measurement in general. So, in this study we will establish detailed flow of high resolution panoramic images and explore spherical panoramic images measured in the forest canopy. Sample area mainly located at the rich forest resources and the diversity of forest—Sitou Nature Park. Sample points will also set upto coniferous, broadleaf, bamboo and park forest, meanwhile we will select different proportion of canopy closure. Eventually, using hemispherical photography is now mature technology shoots at the same sample point. After image analysis software calculating both kinds of images’ canopy parameters and site factors, we compare the difference of panoramic and hemispherical images measured in the forest canopy by inference of one sample mean, two-way analysis of variance and regression analysis. The study results show that both mode of taking photography load into the SOLARCALC7.0 and ArcGIS image analysis software respectively and calculate the value of canopy parameters and site factos by the above statistics. The final results show no significant differences and the correlation is high. It means that high-resolution panoramic images can also be applied to measuring light radiation in the forest. In addition, two kinds of software analyze single shooting mode. The results also shows no significant difference, and the correlation is high. Using images for measuring forest, we recommended high-resolution panoramic image because it still has advantages. High-resolution provide visual analysis, such as habitat destruction and recovery and phenological survey. In future, it can also apply the techniques to positioning and measuring trees, sampling stand and estimating volume and so on. The image analysis software can directly use SOLARCALC7.0.