Abstract | ||
---|---|---|
In this paper, a model-based method is presented for the analysis of SAR images. The problem of image analysis is formulated as the task of cost minimization. Firstly, in the presence of imperfect SAR data, a cost function is defined which combines both of radiometric and geometric properties of objects. Secondly, the estimation procedure is modeled with Markov process, which leads to the use of a globally converged algorithm i.e. the simulated annealing algorithm. Thirdly, the convergence of the algorithm is studied and the order of Markov process is suggested to be used for speeding up the convergence. At last, the experimental data both on testing and real SAR images are provided for evaluating the proposed method. |
Year | DOI | Venue |
---|---|---|
1995 | 10.1007/3-540-60697-1_131 | ICSC |
Keywords | Field | DocType |
sar image,parameter estimation,simulated annealing algorithm,cost function,image analysis,markov process | Simulated annealing,Convergence (routing),Markov process,Imperfect,Experimental data,Pattern recognition,Computer science,Synthetic aperture radar,Minification,Artificial intelligence,Estimation theory | Conference |
Volume | ISSN | ISBN |
1024 | 0302-9743 | 3-540-60697-1 |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
3 |