近年來，有許多研究使用機器學習方法中的監督式分類，搭配空載光達與數位影像資料，進行建物、樹木、道路及草地等地物之詳細分類。然而，建物常與樹木相鄰或重疊，此問題會導致建物與相鄰樹木不易用單一分類方法分離，造成所得建物輪廓有破損情形，在後續提供GIS建物向量圖檔或建物三維建模上無法符合需求。
為了提供較完整的建物輪廓，本研究發展一套「自動化偵測建物被樹木遮蔽區」與「自動化線性特徵復原」演算法的流程。首先，整合最大相似分類法及知識庫為基礎的方法，接著，以自動化方式找出被樹木遮蔽的建物區塊，其次，搭配空載光達高度資料，透過Canny邊緣偵測的方式，提取出建物與樹木邊緣，將兩者影像進行交集運算，因建物區塊已將樹木部份去除，故被樹木遮蓋部份會因交集運算產生缺口。由找出建物被樹木遮蔽部份，再使用邊緣線進行向量化與簡約化提取直線，最後，透過「自動化線性特徵復原」重新調整所偵測直線端點之連接順序，且自動偵測被樹木遮蔽建物區塊的直線破損位置，並進一步將破損的直線有效的補償。
由研究結果得知「最大相似分類法加基於知識庫分類結果校正」與「以物件為基礎之分類法」相比較，二者整體精度差異不太，但以二者分類結果進行自動化偵測建物被樹木遮蔽區時，因前者已加入時間差之處理，經由人工判釋其自動偵測成功率為完全相同，但後者偵測成功率只達67.7%，証明本研究提出之分類方法，其結果較「以物件為基礎之分類法」較適用於建物與樹木遮蔽之偵測。運用「自動偵測建物被樹木遮蔽區」演算法，大幅簡化與加速運算流程。利用「自動化線性特徵復原」演算法，可以自動化方式將破損的直線補償，提供完整的建物輪廓。

In recent years, many researches focused on the supervised classification, one of the machine learning methods, using Lidar and remotely sensed image to provide the four buildings, trees, roads, and grass categories of the ground features. However, buildings and trees are usually much closed or overlapped and this problem will lead buildings and nearby trees not easy to classify by single classification approach. The derived building outlines have many cracks which are not satisfactory for the requirement of GIS building vector map or building 3D modeling.
To provide complete building outlines, this study develops an “automatic detection of the overlapped areas of buildings and trees (ADOABT)” algorithm and an “automatic linear feature recovery (ALFR)” approach to connect building outlines consequently. First, this research integrates Maximum Likelihood Classification (MLC) and Knowledge-Based Correction (KBC) to derive buildings and trees classification resultant images. Next, the ADOABT based on “divide and conquer” principle was used to detect the overlapped areas of buildings and trees. Meanwhile, the building and tree edge images were detected using the Canny edge detector based on Lidar height image. Then, the intersection operator was applied to the detected areas and edge images to detect the crack of the building images. Afterward, vectorization and generalization of the intersection resultant images are applied to extract the straight line of the buildings. Finally, the automatic linear feature recovery procedure was performed to compensate the damage straight line effectively.

According to the experiment results, the classification accuracy derived from integrated MLC and KBC classification method and the object-based classification (OBC) are similar. However, when applying the classification results to detect the overlapped areas of building and trees, because MLC and KBC has the procedure for handling temporal inconsistencies, the success rate of automatic detection is totally the same by artificial interpretation; the detection rate for the results of MLC and KBC is 100% whereas the one for the OBC only 67.7%. It can be concluded that the MLC and KBC approach is more suitable for the automatic detection for the overlapped areas of building and trees. Moreover, the ADOABT algorithm simplifies the workflow of the overlapped area detection. According to the result of edge detection and line detection, the Canny detector presents the clearest edge image. The lines extracted by Vectorized and generalization method are superior to the ones derived from Hough transform. The ALFR algorithm offers a way to connect building outline completely.

中文摘要....................................................................................................I
Abstract …………………………………………….………………….II
誌 謝 …………………………………………….…………………IV
目 錄 ………………………………………….………………….…V
圖目錄 ………………………………………….…………………..VII
表目錄 ……………………………………………….……………….X
第一章 前言 1
1-1 研究動機與目的 1
1-2 文獻回顧 2
1-3 研究構想與流程 7
第二章 影像分類 10
2-1 特徵向量 11
2-2 最大相似分類法 21
2-2-1 實驗區樣本之分離度分析 21
2-2-2 實驗區訓練與測試樣本之選取 24
2-3 基於知識庫分類結果校正 25
2-3-1 形態學斷開 27
2-3-2 高度分析 29
2-3-3 面積分析 29
2-3-4 形狀分析 30
2-3-5 時間差分析 31
2-4 分類精度評估 36
第三章 建物被樹木遮蔽之自動偵測與復原 39
3-1 連通元件分析 43
3-2 自動偵測建物被樹木遮蔽區 44
3-3 邊緣萃取 49
3-4 直線偵測 54
3-4-1 向量化(Vectorize) 54
3-4-2 簡約化(Generalization) 56
3-5 自動化線性特徵復原 60
第四章 研究成果與討論 73
4-1 實驗區資訊 73
4-2 影像分類成果與討論 76
4-2-1 運用光達高度探討以像元為基礎的分類成果之改善程度 77
4-2-2 「以物件為基礎之分類法」之分類成果 79
4-2-3 「最大相似分類法加KBC」之分類成果 81
4-2-4 分類精度評估 89
4-3 建物被樹木遮蔽之自動偵測與復原 92
4-3-1 自動偵測建物被樹木遮蔽之實驗成果 92
4-3-2 邊緣偵測與直線偵測成果 97
4-3-3 自動化線性特徵復原成果 101
第五章 結論與建議 105
5-1 結論 105
5-2 建議 106
參考文獻 ……………………………………………………………...107

1. Alharthy, A., and Bethel, J. (2002). Heuristic filtering and 3d feature extraction from lidar data. In ISPRS Commission III, Symposium 2002 September 9-13, 2002, Graz, Austria Pages A-29 ff (6 pages).
2. Arefi, H., Hahn, M., and Lindenberger, J. (2003). Lidar data classification with remote Sensing tools. Proceedings of the ISPRS Commission IV Joint Workshop Challenges in Geospatial Analysis, Integration and Visualization II, September 8-9, 2003, Stuttgart, Germany, pp 131-136.
3. Baatz, M., Benz, U., Dehghani, S., Heynen, M., Holtje, A., Hofmann, P., Lingenfelder, I., Mimler, M., Sohlbach, M., and Weber, M. (2004). eCognition Professional User Guide 4 ,Definiens Imagine GmbH, Munchen, Germany.
4. Blaschke, T., and Strobl, J. (2003). A comparison of aerial photos, lidar and IKONOS for monitoring cities. Remotely Sensed Cities (Mesev, V., editor). Taylor & Francis, New York, pp. 47-84.
5. Canny, J. (1986). A computational approach to edge detection (IEEE Computer Society Washington, DC, USA), pp. 679-698.
6. Charaniya, A.P., Manduchi, R., and Lodha, S.K. (2004). Supervised Parametric Classification of Aerial LiDAR Data. Conference on Computer Vision and Pattern Recognition Workshop Vol. 3, Washington DC, June 2004.
7. Congalton, R. (1991). A review of assessing the accuracy of classifications of remotely sensed data. Remote Sensing of Environment, pp. 35-46.
8. Davis, S.M., Landgrebe, D.A., Phillips, T.L., Swain, P.H., Hoffer, R.M., Lindenlaub, J.C., and Silva, L.F. (1978). Remote sensing: The quantitative approach.
9. Deering, D. W. & Rouse, J. W. (1975). Measuring forage production of grazing units from Landsat MSS data, Proc., 10th International Symposium on Remote Sensing of Environment ERIM 2, pp. 1169-1178.
10. Douglas, D.H., and Peucker, T.K. (1973). Algorithms for the Reduction of the Number of Points Required to Represent a Digitized Line or its Caricature (University of Ottowa/Simon Fraser University).
11. Haala, N., and Brenner, C. (1999). Extraction of buildings and trees in urban environments. ISPRS Journal of Photogrammetry and Remote Sensing 54, 130-137.
12. Hill, R.A. (2002). Landscape modelling using integrated airborne multi-spectral and laser scanning data. International Journal of Remote Sensing, 23(11): 2327–2334.
13. Holmgren, J., and Persson, A. (2004). Identifying species of individual trees using airborne laser scanner (Elsevier Science Inc. New York, USA), pp. 415-423.
14. Huang, M.-J., Shyue, S.-W., Lee, L.-H., and Kao, C.-C. (2007). A Knowledge-Based Approach to Urban-feature Classification Using Aerial Imagery with Lidar Data. . Photogrammetric Engineering & Remote Sensing (In Press).
15. Hung, M., and Ridd, M.K. (2002). A subpixel classifier for urban land-cover mapping based on a maximum-likelihood approach and expert system rules. Photogrammetric Engineering and Remote Sensing, 68(11): 1173-1180 68.
16. Jain, R., Kasturi, R., and Schunck, B.G. (1995). Machine vision (McGraw-Hill, Inc. New York, NY, USA).
17. Jensen, J.R. (2005). Remote Sensing of Environment: An Earth Resource Perspective,Prentice-Hall, Inc., 2nd Edition.
18. Jiang, B. (2004). Extraction of Spatial Objects from Laser-scanning Data Using a Clustering Technique. XXth ISPRS Congress, Geo-Imagery Bridging Continents, in Istanbul, Turkey.
19. Kovesi, P.D. (2000). Matlab and octave functions for computer vision and image processing.
20. Maas, H.G. (1999). The potential of height texture measures for the segmentation of airborne laserscanner data. Presented at the Fourth International Airborne Remote Sensing Conference and Exhibition / 21st Canadian Symposium on Remote Sensing, Ottawa, Ontario, Canada.
21. Otsu, N. (1979). A Threshold Selection Method from Gray-Level Histograms, IEEE Trans. Systems, Man, and Cyernetices. Vol. SMC-9, no. 1, pp. 62-66.
22. Repaka, S.R., Truax, D.D., Kolstad, E., and O’Hara, C.G. (2004). Comparing Spectral and Object Based Approaches for Classification and Transportation Feature Extraction from High Resolution Multispectral Imagery. ASPRS Annual Conference Proceedings 23-28 May 2004, Denver, Colorado, pp. 204-215.
23. Rottensteiner, F., and Briese, C. (2002). A new method for building extraction in urban areas from high-resolution LIDAR data. In International Archives of Photogrammetry and Remote.
24. Rottensteiner, F., Trinder, J., Clode, S., and Kubik, K. (2003). Building dectection using LIDAR data and multispectral images, in Proceedings of DICTA, Sydney, Australia, pp. 673–682.
25. Song, J.H., Han, S.H., Yu, K., and Kim, Y.I. (2002). Assessing the possibility of land-cover classification using LiDAR intensity data. IAPRS, 9-13 September, Graz, vol 34, 4 p.
26. St-Onge, B.A., and Achaichia, N. (2001). Measuring forest canopy height using a combination of lidar and aerial photography data, pp. 131–137.
27. Sun, W., Heidt, V., Gong, P., and Xu, G. (2003). Information Fusion for Rural Land-Use Classification with High-Resolution Satellite Imagery. IEEE Transactions on Geoscience and Remote Sensing, 41(4): 883-890.
28. Syed, S., Dare, P., and Jones, S. (2005). Automatic classification of land cover features with high resolution imagery and LIDAR data: an object-oriented approach. Proceedings of SSC2005 Spatial Intelligence, Innovation and Praxis: The national biennial Conference of the Spatial Sciences Institute, September, 2005.
29. Taubenbock, H., Esch, T., and Roth, A. (2006). An Urban Classification Approach Based on an Object–Oriented Analysis of High Resolution Satellite Imagery for a Spatial Structuring Within Urban Areas. 1st EARSeL Workshop of the SIG Urban Remote Sensing Humboldt-Universität zu Berlin, 2-3 March 2006.
30. Weidner, U., and Forstner, W. (1995). Towards automatic building extraction from high resolution digital elevation models. ISPRS Journal of Photogrammetry and Remote Sensing, vol 50, pp38-49 50, 38-49.
31. 林承毅 (2004). 空載雷射掃描資料產生數值高程模型之研究， 國立高雄應用科技大學土木工程與防災科技研究所碩士論文.
32. 賴彥中 (2004). 結合光達資料與數位空照影像重建三維建物模型，國立中央大學土木工程研究所碩士論文.