水平樣點取樣及林木上部直徑測量之手機軟體開發

Abstract

為降低水平樣點取樣法與林木上部直徑測量的儀器成本，並於地勢陡峭之臺灣森林進行坡度校正時能有更便捷與準確之儀器，本研究開發手機軟體，將傳統儀器Relaskop之功能以應用程式（Application, APP）實現。研究先瞭解常見儀器於水平樣點取樣法與林木上部直徑測量之原理、功能與優缺點，將所用之幾何方法轉換成數學公式，亦比較市面上森林測計APP之功能與優缺點，以及是否附有坡度校正之功能。接著以Kotlin程式語言進行Android作業系統之APP開發，將數學公式轉為程式碼，開發介面與物件導向程式，將水平樣點取樣法之胸高斷面積指數（Basal Area Factor, BAF）轉換為手機像素，提供坡度與俯仰角校正，確保BAF於不同俯仰角之正確性，實現以手機APP進行水平樣點取樣法與林木上部直徑測量之功能。最後本研究比較所開發之APP於不同手機硬體設備之功能性，選用兩支不同硬體規格之手機（Samsung A22 5G及Redmi7），以傳統Angle Gauge儀器為標準，分別於平地30棵樣樹與坡地15棵樣樹進行BAF為2、4、6、8、10、12、14、16、18、20之實測，分析邊界距離（Limiting Distance, LD）之測量值與理論值，以比較APP於不同手機與傳統儀器Angle Gauge之準確度，特別是在手機前鏡頭解析度與螢幕解析度不同時對APP表現之影響。

分析結果發現，在平地測試中，APP於Samsung手機之平均誤差為-4% ~ -5%，於Redmi手機之平均誤差為-3% ~ -5%，而Angle Gauge之平均誤差則為-1% ~ 6.5%；坡地測試中，APP於Samsung手機之平均誤差為-5% ~ -6%，於Redmi手機之平均誤差為-4% ~ -5%，而Angle Gauge之平均誤差則為1.5% ~ -2.5%。由實驗結果推論，手機APP於不同硬體之準確度相似，但Angle Gauge之準確度仍較手機APP高。誤差分佈分析結果顯示手機APP之量測精度較高，Angle Gauge之量測結果則有較多變異。因此本研究所開發之手機APP具有良好之準確度，雖平均誤差較傳統儀器高，但具有較佳之量測精度，且不易受硬體規格影響，具有取代傳統高成本儀器，成為新型態森林測計工具之潛力。

To reduce the cost associated with Horizontal Point Sampling (HPS) equipment and upper-stem diameter measurements, and to enhance measurement accuracy in Taiwan’s steep terrains by slope correction, this research develops a smartphone application (APP) that emulates the functions of the classical Relaskop instrument and includes key ability to correct for slope.

Firstly, the research conducts a comparative analysis of various equipment types for HPS and upper-stem diameter measurements to grasp their principles, functions, and advantages and disadvantages, translating the principles into mathematical formulas as well as developing advanced geometric methods. Subsequently, the research assesses different forest mensuration applications available on Google Play, examining whether they offer slope correction functionality.

Furthermore, the research employs the Kotlin, an object-oriented programming language, to develop an Android smartphone APP. This APP calculates the relative angle pixels of Basal Area Factor (BAF) captured by the smartphone camera and employs geometric method to measure upper-stem diameter. To ensure the correctness of the BAF under different slopes, the APP utilizes the mathematical formulas to conduct slope correction.

To evaluate the general performance of this smartphone APP across different hardware devices, the research compares its accuracy and precision on Samsung A22 5G and Redmi7 with that of the traditional equipment, the Angle Gauge. The study evaluates the APP using BAF values ranging from 2 to 20, analyzing 30 sample trees on a flat terrain plot and 15 on a steep terrain plot. The research analyzes the error between the measured Limiting Distance and the theoretical Limiting Distance to compare the accuracy of the APP on different smartphones and the Angle Gauge.

The experiment results indicate that, for flat terrain plots, the App on Samsung and Redmi smartphones yields mean errors ranging from -4% to -5% and -3% to -5%, respectively, and the Angle Gauge exhibits errors ranging from -1% to 6.5%, indicating greater variability. Similarly, for steep terrain plots, the smartphone APP on Samsung and Redmi devices shows mean errors of -5% to -6% and -4% to -5%, respectively, while the Angle Gauge demonstrates errors from 1.5% to -2.5%. From these results, the APP on smartphones has similar accuracy across different devices, although the Angle Gauge has a higher accuracy than the APP. The APP has a higher measurement precision, while the results of the Angle Gauge are more variable. Based on the above, the research concludes that the APP have a fair accuracy. Although the mean error of the APP is higher than the Angle Gauge, the APP shows a better precision, and the performance would not be affected by the hardware specifications, demonstrating the potential to replace traditional high-cost instruments as new type of forestry measurement tool.