請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/24061
標題: | 超音波斷面影像技術應用於立木缺陷之評估 Defect Evaluation of Standing Trees by Ultrasonic Tomography Technique |
作者: | Yu-Chien Kao 高毓謙 |
指導教授: | 蔡明哲(Ming-Jer Tsai) |
關鍵字: | 多路徑超音波斷面影像技術,人造孔洞,傳遞時間, Ultrasonic tomography technique,Artificial hole,Propagation time, |
出版年 : | 2006 |
學位: | 碩士 |
摘要: | 本研究應用多路徑超音波斷面影像技術,評估柳杉圓盤內部之不同大小中央及偏心人造孔洞,以及腐朽之立木,以瞭解多路徑超音波斷面影像技術之適用性。研究中使用SYLVATEST Duo®超音波儀檢測,以人工逐次挖鑿柳杉圓盤中央及偏心孔洞模擬不同心腐程度,探討不同路徑超音波速度及相對速度損失率與孔洞直徑關係,及使用超音波傳遞時間比值和2D斷面影像圖形平均速度與孔洞直徑比值、孔洞殘餘比值、孔洞面積比值進行討論,及利用不同演算法所得結果差異比較,與使用不同數量的偵測點對於影像解析提升的效果。
實驗數據結果顯示圓盤中央孔洞直徑殘餘比值與超音波速度在Type D(180º)有最高的相關性,其R2值可達0.94,依次為Type C(135 º)>Type B(90 º)>Type A(45 º)。在超音波2D斷面影像圖形中發現能顯示出孔洞位置與大小,在圓盤孔洞部分明顯與周圍顏色有差異存在,亦表示著超音波速度的差異,相對的圓盤材質狀況也有差異,在孔洞/圓盤斷面積比值約2.7%以上時,皆能顯示出圓盤內部孔洞位置,在偏心孔洞上,利用正徑向與正弦向比值評估內部孔洞材質,結果顯示此方法可以正確評估缺點所在的象限,且隨著偏心孔洞愈大,其影像也可以明顯的表示出來,可以供未來長期監測林木內部材質使用。 在不同偵測點評估中發現,偵測點必需大於8個以上,才能有效的預估內部孔洞,在16個與32個偵測點結果中發現影像差異不大,皆能降低因層析成象所造成的干擾,在三個不同演算方式結果中,其使用ARBOTOM的效果為佳,其餘二者差異不大,三者皆無法顯示出「L型」孔洞結果,僅能顯示填補成「凸邊三角形」的影像,與以往之研究比較,已提升評估之準確性,可以應用做為評估林木樹幹心腐程度之參考,配合VTA(Visual Tree Assessment)法可做為林木危險度評估之依據。 The purpose of this study was to investigate decayed standing trees and artificial holes in the cross section of Cryptomeria japonica by using ultrasonic tomography technique. The ultrasonic instrument Sylvatest-Duo and the trunk section of Cryptomeria japonica were used in our study. In order to simulate different degrees of the core decay, chiseling on the central area of trunk section was done by hand and then the ultrasonic velocities of various measured directions were investigated. The relationships among different values of ultrasonic velocities, the relative velocity drop, and the diameter of the chiseled hole were analyzed. Finally, we attempt to conclude from the results of different algorithm and the measured points advancing tomographic technique. Moreover, we discussed the relationships among the ratio of propagation time in the 90° direction to radial (180°) direction, the average velocity of ultrasonic tomography, the ratio of the hole/section diameter, the ratio of the hole remnant, and the ratio of the hole/section area. The highest correlation coefficient was found between the ratio of central hole remnant and the ultrasonic velocity in route D (two adjacent measured point), whose R2 value is up to 0.94. The following order according to correlation coefficient was route C (two-point interval between two measured point), route B (one-point interval between two measured point), and route A (two measured point are adjacent to each other), respectively. It was also found that the location and the size of the hole in the tree trunk section could be detected by the ultrasonic tomography. The color difference between the hole area and its surrounding area was obvious by ultrasonic tomography analysis. When the ratio of hole area to section area was larger than 2.7%, the hole was detectable by using ultrasonic tomography. Besides, we could also find out the precise quadrant in which the flaws were located by applying the method above in centrifugal whole. The ultrasonic image is clearer when the centrifugal hole is larger. This application can do a great favor to the long-term monitoring of inner quality of stems. According to the experimental results, the undergoing technique could efficiently predict inner holes with eight measured points at least, while the ultrasonic image would resemble the actuality and reduce the inference by tomography with 16 and 32 measured points. Analyzing the three results of the different algorithm, we could find that the one which using ARBOTOM is the best, while the other two had little difference, and all of them couldn’t show L-shape holes, but triangles. In comparison with the former studies, this technique has enhanced the accuracy of investigation. As a result, the obtained results could be used for the evaluation of core decay, and could be used as a basis for the visual tree assessment. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/24061 |
全文授權: | 未授權 |
顯示於系所單位: | 森林環境暨資源學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-95-1.pdf 目前未授權公開取用 | 30.3 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。