Implementation of Image Registration for Satellite Images using Mutual Information and Particle Swarm Optimization Techniques
| Autor: | Heena R. Kher |
|---|---|
| Rok vydání: | 2014 |
| Předmět: |
business.industry
Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Particle swarm optimization Image registration Mutual information Image segmentation Swarm intelligence Joint entropy Computer Science::Computer Vision and Pattern Recognition Computer vision Artificial intelligence Multi-swarm optimization business Flocking (texture) Change detection |
| Zdroj: | International Journal of Computer Applications. 97:7-14 |
| ISSN: | 0975-8887 |
| DOI: | 10.5120/16969-5475 |
| Popis: | aim of this research is to register satellite images on the DSP processor using probabilistic optimization method named as particle swarm optimization. Satellite image registration is necessary in order to find change detection, to eliminate influence of camera distortion (roll, pitch and yaw), merge satellite imagery and in urban planning. Particle Swarm Optimization is a stochastic search technique with less computation and still very effective as compared to other optimization techniques. It is based on bird flocking, fish schooling and swarm theory. Each particle changes its position and velocity based on its corresponding fitness value. Fitness value can be calculated using joint entropy and mutual information. The algorithm can be used in object recognition, image segmentation, matching and registration. The performance of this algorithm is measured and results are shown using DSK 6713 hardware along with VM32242. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|