(3.236.214.19) 您好!臺灣時間:2021/05/09 22:59
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

: 
twitterline
研究生:蔡紹安
研究生(外文):Shau-An Tsai
論文名稱:Pollock模型最小平方法之樹冠偵測與描繪
論文名稱(外文):Least Square Fitting of Pollock Model for tree detection and crown delineation
指導教授:吳昭正
指導教授(外文):Chao-Cheng Wu
口試委員:林金樹詹寶珠張建禕
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:電機工程研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
畢業學年度:104
中文關鍵詞:Pollock模型、最小平方法、樹木偵測、樹冠描繪
外文關鍵詞:Pollock model; active contour; tree detection; tree crowns delineation; bottom-up erosion
相關次數:
  • 被引用被引用:0
  • 點閱點閱:73
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:4
  • 收藏至我的研究室書目清單書目收藏:0
近年來氣候越趨近極端,大自然的保護更為重要,如何有效的管理森林面積以及利用樹種變遷來追蹤氣候的變化,成為一個非常重要的議題。但因為透過人工方式實地去調查非常費時且耗人力,由於遙測影像具有大範圍與即時監控等特性,因此該技術被廣泛地用於此一應用。
自1980年以來有許多的樹木偵測與輪廓描繪的演算法被提出,最早的演算法是利用樹木於彩色(RGB)空照圖上的特性,將每棵樹的輪廓圈選出來。近十多年來,由於光達影像可以提供物體垂直的結構,因此開始被廣泛地用於此應用。縱然許多的針對多光譜或RBG彩色影像所提出演算法依然適用於光達影像,然而這些演算法均未善加利用光達影像所提供的樹木垂直特性。
本論文利用Pollock所提出的三維樹木模型,此一模型可以用來模擬樹木三維的輪廓,然而此一模型的精確度嚴重的仰賴各項參數。為了解決此一問題,本論文提出一套迭代式演算法,於迭代的過程中找出與樹冠頂端的資料點最接近的Pollock模型,並透過最小平方法求得誤差,利用此一誤差將樹冠區域過濾出來,並將其概略的輪廓描繪出來,所提出的演算法將提供更精確的樹木偵測率與快速的運算,使其更適合運用於樹木的偵測與描繪。
In Taiwan, the forests distribute vertically along the central region and can be categorized into hardwood, mixed, and conifer forests. Terrain features in mountainous areas make manual inspection of forests very difficult if not impossible. By utilizing remote sensing techniques the efforts of field sampling could be effectively reduced.
The related remote sensing research was beginning from digital imagery back to the mid-1980s. One of the earliest work of such algorithms presented by Pinz proposed the vision expert system to locate the center of a crown and estimate the crown radius by searching for local brightness maxima in smoothed aerial images. Afterwards many algorithms had been proposed and developed. In recent years, Light Detection And Ranging (LiDAR) data have emerged as sources for estimating tree height, stem volume and biomass at stand level due to strengths of three dimensional structures. High sampling LiDAR data provides detailed vertical structure of tree crowns. Although some algorithms developed for high-resolution optical imagery could be applied for extracting individual tree information from LiDAR data, few of them were specifically designed to utilize 3D properties provided by LiDAR data.
The template matching algorithms shed light on ideas of the algorithm proposed in this thesis. The templates are constructed by considering a three-dimensional description of individual tree-crown envelope. Pollock proposed a well-known model to construct a synthetic comprehensive image template considering both geometric and radiometric characteristics of individual tree. The Pollock model helps to increase detection rate if parameters of models were set properly. Nevertheless, it is always challenging to determine parameters since each individual tree has different grown condition. To address this problem, priori knowledge about the scene is required to predetermine an appropriate range of parameters for the models.
This thesis proposed an iterative algorithm, called Least Square Fitting of Pollock Model. It designs an iterative process to look for the Pollock model best fitting the real 3D structure of each extracted tree top. The proposed algorithm provide a solution to avoid difficulties of finding the best set of parameters for each tree crown. As a result, it improves the detection rate and decreases the computing time as shown in the experimental studies. The proposed algorithm is more suitable for practical applications.
摘 要 I
Abstract II
誌 謝 IV
Contents I
List of Tables III
List of Figures IV
Chapter 1 INTRODUCTION 1
1.1 Motivation and Background 1
1.2 Introduction 1
Chapter 2 BACKGROUND 3
2.1 Overview 3
2.2 Multi-level Morphological Active Contour (MMAC) 3
2.2.1 Bottom-up erosion (BUE) 4
2.2.2 Top-down dilation (TDD) 5
2.2.3 Active contour model (ACM) 6
2.3 Pollock Model 7
Chapter 3 RESEARCH METHODS 11
3.1 Research framework 11
3.2 Bottom-up erosion modifying (BUE) 12
3.3 Pollock model algorithm 12
3.4 Intersection of ratio 17
3.5 Analysis Introduction 19
Chapter 4 EXPERIMENT RESULTS 22
4.1 Image Information 22
4.2 Simulation Images 23
4.3 The analysis of real image 28
Chapter 5 CONCLUSION AND FUTURE WORK 40
5.1 Conclusion 40
5.2 Future work 41
REFERENCE 42
1.Lin C., Thomson G., Lo C.S., Yang M.S., “A multi-level morphological active contour algorithm for delineating tree crowns in mountainous forest,” Photogrammetric Engineering and Remote Sensing 77(3): 241-249, 2011.
2.Erikson, M., and K. Olofsson, 2005. Comparison of three individual tree crown detection methods, Machine Vision and Applications, 16(4):258–265.
3.Kass, M., A. Witkin, and D. Terzopoilos, 1988. Snakes: Active contour models, International Journal of Computer Vision,1(4):321–331
4.Yinghai Ke; Lindi J. Quackenbush, A review of methods for automatic individual tree-crown detection and delineation from passive remote sensing. International Journal of Remote Sensing, 2011, 32:17, pp. 4725-4747, DOI:10.1080/01431161.2010.494184.
5.Pollock, R., The automatic recognition of individual trees in aerial images of forests based on a synthetic tree crown model, Ph.D. dissertation, Department of Computer Science, University of British Columbia, Vancouver, BC, 1996.
6.Pitkänen J., Maltamo M., Hyyppa J. and Yu X., “Adaptive methods for individual tree detection on airborne laser based canopy height model,” ISPRS: Laser-scanners for forest and landscape assessment, 3-6 October 2004, Freiburg, Germany (ISPRS), pp. 187-191, 2004
7.Gong P, Sheng Y, Biging GS. 2002. 3D model-based tree measurement from high-resolution aerial imagery. Photogrammetric Engineering and Remote Sensing 68(11): 1203-1212 .
8.Pollock, R.J.: A model-based approach to automatically locating tree crowns in high spatial resolution images. Image Signal Process. Remote Sens. (eds.) Desachy, SPIE, vol. 2315, pp. 526–537 (1994)
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關論文
 
無相關期刊
 
無相關點閱論文
 
系統版面圖檔 系統版面圖檔