164 
Data 
Method 1 
Method 2 
Method 3 
time(s) 
Error (pixel) 
time (s) 
Error(pixel) 
time (s) 
Error(pixel) 
Al-Bl 
216. 23 
0. 566 
20. 92 
0. 933 
21.21 
0. 570 
A2-B2 
198.50 
0. 521 
19. 68 
0. 878 
19. 87 
0. 536 
A3-B3 
221. 23 
0. 535 
21.25 
0. 857 
20.86 
0. 520 
A4-B4 
232. 13 
0. 580 
23.43 
0. 951 
23. 51 
0. 576 
Table 1 Comparison of registration precision and time 
7. CONCLUSIONS 
In this paper, an novel high-precision registration algorithm for 
images with different time, sensor and resolution is proposed 
and the result of experiment using TM and SPOT images 
showed that: 
1) The method proposed in this paper has good precision and 
low time cost. The grid technology based on entropy could 
make the extracted feature points distribute evenly. 
2) Based on a few of accurate control points, the registration 
algorithm using invariant moments similarity measurement 
combined with global relaxation matching has high tolerance 
degree to geometry transformation and gray difference. 
3) Eliminating mismatch points using quadratic polynomial 
model is feasible and could improve the precision of 
registration. 
REFERENCES 
[1] Strunz G. Automation of the Pointing Procedure for the 
Geocoding of Satellite Images. International Archives of ISPRS, 
1994(30):793-800. 
[2] Djam dji J. Geometrical Registration of Image s: the 
Multiresolution Approach Photogrammetry and Remote 
Sensing, 1993, 59(5):645-653. 
[3] M K Hu. Visual Pattern Recognition by Moments Invariant 
[J], IRE Trans, inf. Theory 8, 1962, 179-187. 
[4] ZHANG Zu-xun, ZHANG Jian-qing, LIAO Ming-sheng etal. 
Automatic Precision Registration Multi-resolution Remote 
Sensing Imagery [J], Journal of Wuhan Technical University of 
Surveying and Mapping, 1998.23(4):320-323. 
[5] ZHENG Ming-ling, LIU Heng-zhu. High Performance 
Algorithm of Selecting Characteristic Points in the Remote 
Sensing Images Registration and Its Implementation [J],Journal 
of Computer, 2004.27(9): 1284-1289. 
[6] ZHANG Jix-ian, LI Guo-sheng, ZENG Yu. The Study on 
Automatic and High-Precision Rectification and Registration of 
Multi-source Remote Sensing Imagery [J].Journal of Remote 
Sensing, 2005.9(l):73-77. 
[7] SUN Jun-ding, WU Xiao-sheng, ZHOU Li-hua. Entropy- 
based Image Retrieval [J].Joumal of XIDIAN 
University ,2004.31(2):223-228. 
[8] ZHANG Zu-xun, ZHANG Jian-qing. Digital 
Photogrammetry[M]. Wuhan : Wuhan Technical University of 
Surveying and Mapping Press, 1997. 
[9] YAN Bo-jun, ZHENG Lian, WANG Ke-yong. Fast Target- 
Detecting Algorithm Based Invariant Moment[j]. Infrared 
Technology, 2001, 11(6):8-12. 
[10] SHEN Bang-le. Computer Images Processing[M] .PLA 
Press, 1995: 93-128. 
[11] Kenneth R. Castleman. Digital Image Processing 
[M] .Beijing: Publishing House of Electronics Industry, 2002.2: 
100-127. 
[12] FANG Yong, CHANG Ben-yi.The Research of Removing 
the Affect of Clouds Cover by Combining the Multi-sensor 
Images[J].Joumal of Image and Graphics, 2001.6(2): 138-141. 
[13] LI Xin-tao. Research of Information Updating technology 
of Digital Remote Sensing Images [D],Zhengzhou: PLA 
University of Information and Engineering Master Paper, 2005. 
[14] CHEN Yin. Digital Photogrammetry on Remote Sensing 
Images [М].Shanghai: Tongji University Press, 2003.9:102-105