pt U2 Buda VE US 
«D £A 
  
  
  
  
unmosaicing result of two kilometer grids. 
Fig.18 shows the mosaicing result 
corresponding to Fig. 17. 
A test production for land source 
investigation is made using many pieces of 
land use map. These maps are processed to 
test the accuracy of area and length 
measuring. Comparing with the chinese 
land source investigation standards, the 
qualified ratio reaches to 100 percent. 
And this is not the highest accuracy of CV2, 
Reducing the working patch can increase the 
accuracy when higher precision is expecting. 
A set of experiments show that the scheme 
of the thematic map reading system 
proposed in this paper is realizable, 
Optimal combination of computer —visiov 
techniques and manual operation raises the 
global efficiency of the system, Multi 
-criteria line following improves the 
reliability and adaptibility of line 
extration, Topological data structure 
organizes the thematic map data 
efficiently and makes it easy to store and 
manage them, Using this system to land 
source investigation has get to a practical 
stage. On the other hand, the idea and 
some algorithms are suitable for the 
reading process of other line drawings. So 
this system can be developed further. 
REFERENCES : 
[1] Ballard, D.H. & Brown, C M , 1982. 
Computer vision, Prentice Hall. 
[2] Beard, M.K. & Chrisman, N. R. , 1986. 
Zipping: new software for merging map 
sheets, technical papers of 1986 ACSM 
-ASPRS annual convention, Vol.1. 
[3] Cui,Weiping, 1989.Design and research of 
multipurpose cadastre information system, 
Journal of Zhengzhou institute of 
surveying and mapping, No.2, in chinese, 
[4] Fishler, M. A. , Tenenbaum, J.M & Wolf, H.C., 
1981. Detection of roads and linear 
structures in low resolution aerial 
imagery using a multi- source knowledge 
integration technique, CGIP, Vol. 15, 
201-223. 
[6] Groch, W.D., 1982. Extraction of line 
shaped objects from arerial images 
using a special operator to analyse the 
profile of functions, CVIP, Vol. 18, 
347-368. 
[6] Hung, S.H.Y & Kasvand, T. , 1983. 
Critical points on a perfectly 8-or 6- 
connected thin binary line, 
Pattern 
243 
[7] 
[8] 
[9] 
[10] 
[11] 
[12] 
[13] 
[14] 
[16] 
[16] 
[17] 
[18] 
Recognition, Vol. 16, No. 3, 297-306. 
Joseph, S. H. , 1989. Processing of 
engineering line drawings for automatic 
input to CAD, Pattern Recognition, 
Vol. 22, No.1, pp 1-11. 
Kasturi, R. etal., 1990. A system for 
interpretation of line drawings, IEEE 
Trans, on PAMI, Vol. 12, No. 10. 
Lin Zongjian, Lu Jian & Fu Zhongliang, 
1990a. GIS data inputing through 
CCD camera, ISPRS, vol.28-3/2, May, 1990. 
Lin Zongjian, Lu Jian & Yan Li, 1990b. A 
topographic contour lines reading 
system, Proc. of the 11th Asian conf. on 
remote sensing, Guangzhou, China, NOV. 
Lin Zongjian & Lu Jian , 1991a.  CV2 
graphic reading system , technical 
report , Wuhan technical university of 
sureying and mappimg, May, in chinese, 
Lin Zongjian, Lu Jian, Yu Zifan & Di 
Kaichang, 1991b. Computer vision 
based map reading system, FIG int. symp., 
Beijing, May. 
Musavi, M. T. , Shirvaikar, M. V., Ramanathan, 
E & Nekovei, A.R.,1988. À vision based 
method to automatic map processing, 
Pattern Recognition, Vol. 21, No.4, pp 
319-326. 
Nagasamy, V & Langrana, N. A , 1990. 
Engineering drawing processing and 
vectorization system, CVGIP, Vol. 49, 379 
-397. 
Shu, Joseph Shou-Pyng, 1989. One- pixel 
-wide edge detection, Pattern Recognition, 
Vol. 22, No.6, pp 666-673. 
Suzuki, S. & Yamada, T., 1990. MARIS 
map recognition input system, Pattern 
Recognition, Vol. 23, No.8, pp 919-933. 
Zamperoni, P., 1982. Contour tracing of 
grey scale images based on 2-D histograms, 
Pattern Recognition, Vol.16, No. 3, pp 
161-166. 
Zhu, Zhipu & Kim Yongmin, 1989. 
Algorithm for automatic road recognition 
on digitized map images, Opt. Eng. , Vol. 
28, No. 9.