120 
have been discarded in Figure 8. However, the detection result 
shows degraded edge connectivity, this can be made up by 
some following processing such as edge linking. The ROI 
boundary image which is supposed to have better connectivity 
and less fake edges is not regarded as the final detection result, 
because the location of the edge is no longer accurate after 
image segmentation and a series of filtering process. 
Figure 8 Optical Canny Edges Figure 9 Detection Result 
5. CONCLUSION 
A detection method for line-type targets based on SAR and 
optical image fusion is proposed in this paper. The region 
integrality of SAR images and the legible edge of optical 
images are utilized in the method. It can be easily realized, 
since the involved image processing technologies are all well- 
developed and mature. Experiment results demonstrate the 
feasibility and the validity of the proposed method. However, 
some improvement can be made in the following aspects. The 
final detection result should be provided with better edge 
connectivity on the premise of algorithm simplicity and 
speediness; different pixel location should be given a different 
believe factor to generate a more better result. 
REFRENCES 
Chee Sun Won and Haluk Derin, 1987. Segmentation of Noisy 
Textured Images Using Simulated Annealing. IEEE Proceeding 
of 1987 International Conference on ICASSP, pp.563-566. 
E.Lallier, M.Farooq, 2000. A Real Time Pixel-level Based 
Image Fusion Via Adaptive Weight Averaging. Proceedings of 
the Third International Conference on Information Fusion, 
Vol.2, pp. 10-13 
Li Ming, Wu Yan, Wu Shunjun, 2004. A New Pixel-level 
Multi-focus Image Fusion Algorithm Based on Evolutionary 
Strategy. Control, Automation, Robotics and Vision Conference 
, Vol.2, pp. 810-814 
Min-Sil Yang, Wooil M.Moon, 2003. Decision Level Fusion of 
Multi-Frequency Polarimetric SAR and Optical Data with 
Dempster-Shafer Evidence Theory. Geoscience and Remote 
Sensing Symposium, IGARSS Proceedings, Vol.6, pp. 3668- 
3670 
Robert A., Weisenseel W., Clem K., et al, 1998. MRF-Based 
Algorithms for Segmentation of SAR Images. IEEE Proceeding 
of the 1998 International Conference on Image Processing, 
Paris, pp.770-774 
S. Geman and D.Geman. Stochastic relaxation, Gibbs 
distribution and Bayesian restoration of image. IEEE Trans. 
Pattern Analysis and Machine Intelligence, 1984. 6: 721-741.