在建物模型的自動化萃取方法中，常以建物邊緣線的搜尋為出發。進行建物邊緣線的自動化偵測和萃取時，由於一般灰值邊緣線的萃取方式只能顧及影像中灰值的變化以及像空間內二維幾何特徵物，而又受到影像中雜訊、陰影、遮蔽、弱反差或是建物本身紋理的影響，以致於低階萃取所得的邊緣線常是破碎、不連續的，而且含有許多非建物邊緣線。由於萃取時並未考慮物空間三維資訊，以致於所得到的影像中邊緣線大部分與實際物空間邊緣線有很大的差距。
本研究以物空間的觀點，分析直線段在物空間和影像空間中的特性，藉由一個適合多張影像中邊緣線在物空間中匹配成直線的函數模式，嘗試將影像萃取出來的低階直線型特徵物(linear feature)匹配成在物空間內有意義的三維直線段，並以統計分析方法找出當匹配為正確時，各像元殘差應具有的統計特性，以作為判斷匹配正確與否的依據。經過統計檢定便能找出在物空間內可匹配成直線段的線元(line element)。
結果顯示以空間直線函數模式做為約制，來尋找影像中可能在物空間中組成三維直線段的對象，確實可以顧及物空間的觀點篩選出正確匹配者。而本方法亦可在匹配的過程中同時進行鏈結，可以由破碎的線段中獲得更長、更完整的線段，提供較完美的三維線段供後續建物模型的重建之用。

Automatic extraction of building models is proceeded often by searching building edges which in turn are based on low level edge extraction in images. But low level edge extraction from gray value considers only variation in gray value and some 2D geometric properties in image space. Moreover, it is also affected by noises, shadows, occlusion and wake contrasts in images or even the building texture itself. As a result, edges of low level extraction are usually not only broken, segmental but also contain many non-building edges as well. The result is that most of the edges in images do not correspond to actual edges in object space.
This research analyzed the properties of line segments both in object space and in image space from the perspective of object space. By using a function model for fitting edges in images onto a straight line in object space, we could construct meaningful 3D line segments in object space by simultaneously matching of low level linear features extracted from multiple images. Meanwhile, statistical properties of residual of image coordinate after correct matching were analyzed and statistical test values were derived for testing the correction of matching. By doing these tests, those low level edges in image space could be selected out which can form straight line segments in the object space.
Results show that object space knowledge is effectively taken into consideration by using straight line function model as an object space constraint to find possible targets of forming 3D straight line segment in images is able to give consideration to the viewpoint of object space as well. In addition, the function model could also serve as a tool for linking broken 2D gray value edges simultaneously with matching, then helping to acquire longer and more complete 3D line segments for the subsequent reconstructions of building models.

邱式鴻、王蜀嘉「Förstner特徵物萃取法精確性的探討」，第十五屆測量學術及應用研討會，p667-676，1996。
邱式鴻、王蜀嘉，「以立體像對為基礎的線型特徵物鏈結方式」，第十六屆測量學術及應用研討會，p665-674，1997。
邱式鴻，「從都市區立體航照影像中萃取屋頂面的實用策略」，國立成功大學測量工程研究所博士論文，2000。
吳碧玉，「以二維CAD數值圖輔助屋頂模型之重建」，國立成功大學測量工程研究所碩士論文，1997。
李德仁，「誤差處理與可靠度理論」，測繪出版社，1988。
張祖勛、張劍清、梅雪良、朱英浩、曹輝，「航空影像規則房屋自動三維重建」，第二屆兩岸測繪學術研討會暨第十七屆測量學術及應用研討會，p748-756，1998。
陳良建、陳慶明鳥、饒見有，「數位空照立體對邊緣萃取與匹配之研究」，第十五屆測量學術及應用研討會，p687-696，1996。
陳世崇，「以屋頂面法線為約制條件協助立體像對中建物邊緣線之篩選」，國立成功大學測量工程研究所碩士論文，1998。
Baillard, C., Schmid, C., Zisserman, A. and Fitzgibbon, A., "Automatic Line Matching and 3D Reconstruction of Buildings from Multiple Views", International Archives of Photogrammetry and Remote Sensing, vol.32, pp.69-80, 1999.
Baillard, C., and Dissard, O., "A Stereo Matching Algorithm for Urban Digital Elevation Models", Potogrammetic Engineering and Remote Sensing, pp.1119-1128, 2000.
Bignone, F., "Segment Stereo-Matching and Coplanar Grouping", Institute of communication Technology. Image Science Lab, ETH, Zurich, Switzerland, 1995.
Chio, S.-H. and Wang, S.-C., "Semi-Automatic System for Roof Reconstruction Based on 3D Linear Segments", the 20th Asian Conference on Remote Sensing,vol.1,pp.165-170,1999.
Chio, S.-H., Wang, S.-C. and Wrobel, B., "Interactive Roof Patch Reconstruction Based on 3-D Linear Segments", International Archives of Photogrammetry and Remote Sensing, vol. XXXIII pp.177-184,2000.
Elaksher, A.F., Bethel, J. S. and Mikhail, E.M., ”Building Extraction Using Multiple Images”, ACSM-ASPRS 2002 Annual Conference Proceedings, 2002.
Förstner, W., "A Framework for Low Level Feature Extraction", Ekiundh, J.-O.(Ed.), Computer Vision, ECCV'94, vol.ll,pp.383-394, Lecture Notes in Computer Science, 801, Springer-Verlag, Berlin, 1994.
Förstner, W., "Geometric and Statistical Models for Image Analysis".
Fuchs, C., "Feature Extraction", Förstner, W. and Winters, S., editors, Proc. of Second Course in Digital Photogrammetry, 1995.
Strunz, G., ”Bildorientierung und Objektrekonstruktion mit Punkten,Linien und Flächen”, Dissertation, Deutsche Geodätische Kommission, Verlag der Bayerischen Akademie der Wissenschatten,München, 1993.
Haala, N., "Detection of Buildings by Fusion of Range and Image Data", International Archives of Photogrammetry and Remote Sensing vol.30, pp.341-346, 1994.
Haala, N., "3D Building Reconstruction using Linear Edge Segment", Fritsch, D., Hobbie, D., Fritsch, D. and Hobbies, D., editors, Photogrammetric Week’95, pp.1-28, Herbert Wichmann Verlag, Heidelberg, 1995.
Henricsson, O. and Baltsavias, E., "3-D Building Reconstruction with ARUBA: A Qualitative and Quantitative Evaluation", Gruen, A., Baltsavias, E. P. and Henricsson, 0., Gruen, A., Baltsavias, E. P. and Henricssons, 0., editors, Automatic Extraction of Man-Made Objects from Aerial and Space Images (II), pp.65-76, Birkhäuser Verlag,1997.
Park, H. J. and Zinmmermann, P., "Colour Image Matching For DTM Generation And House Extraction", International Archives of Photogrammetry and Remote Sensing, Vol. XXXIII, Part B3.Amsterdam,pp.697-704,2000.
Medioni, G. and Nevatia, R., "Segment-Based Stereo Matching", Computer Vision, Graphics, and Image Processing (31), pp.2-18, 1985.
Moons, T., Frére, D., Vandekerckhove, J. and Van Gool, L., "Automatic Modelling and 3D Reconstruction of Urban House Roofs from High Resolution Aerial Imagery ", Computer Vision-ECCV'98: 5th European Conference on Computer Vision, vol.I,pp.410-425, Springer Verlag,1998.
Luong, Q.-T. & Viéville, T. "Canonic Representations for the Geometries of Multiple Projective Views", Lecture Notes in Computer Science,Vol.800, Jan-Olof Eklundh(Ed.) Computer Vision-ECCV’94,Springer Verlag Berlin Heidelberg,1994.
Schmid, C. and Zisserman, A. "Automatic Line Matching across Views",1997 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,pp.666-671,IEEE Computer Society,1997.
Scholze, S., Moons, T., Ade, F. and Van Gool, L., "Exploiting Color For Edge Extraction And line Segment Stereo Matching In High-Resolution Aerial Imagery", International Archives of Photogrammetry and Remote Sensing, vol. XXXIII, Part B3.Amsterdam,pp.815-822,2000.
Hartley, R. and Zisserman, A., "Multiple View Geometry in Computer Vision",2001.
Schenk, T., "Digital Photogrammetry vol. I", TerraScience, 1999.