Segmentation of Synthetic Aperture Radar by Artificial Bee Colony(ABC)

Segmentation of Synthetic Aperture Radar by Artificial Bee Colony(ABC)

anaei seyede maryam1 hematpour naghi2

1) jahad university of yasooj,
2) jahad university of yasooj,

IOI Code : XYAH-ZDAAK
Publication : 3rd international conference of Science and Engineering(3icesconf.com)
Abstract :
    Due to the Presence of speckle noise , segmentation of synthetic aperture radar (SAR) is still an open question. The method for segmentation SAR images segmentation based on artificial bee colony algorithm are (ABC) offers. Estimate the threshold as a search process, to find the appropriate values, Gray scale continuous range to be searched. Therefore, the algorithm ABC, to search the optimal value threshold be used. In algorithm ABC to obtain a proper fitness function, after defining Gray number in theory Gray, the original image is decomposed using the method Violet. Then, by applying noise reduction function is built on the image approximate of the low frequency coefficients, the filtered image is produced. At the same time gradient image is reconstructed with high frequency coefficients. Thus, the Syndicate matrix can be built on the basis the filtered image and gradient,and Gray Improved entropy in two dimensions, as a fitness function algorithm ABC, is defined. Finally, using the worker, observer and scout bees in the bees colony, the optimal threshold is achieved in less time. Experimental results show that the proposed approach rather than genetic algorithms and fish swarm algorithm, in terms of accuracy and time of image segmentation is better.

Keywords : segmentation genetic algorithm fish swarm algorithm artificial bee colony algorithm