TY - JOUR
KW - Image Segmentation
KW - Multilevel Thresholding
KW - Otsu’s Entropy
KW - Electromagnetic Optimization
KW - Levy Function
AU - Ashraf Hemeida
AU - Radwa Mansour
AU - M E Hussein
AB - Image segmentation is considered one of the most important tasks in image processing, which has several applications in different areas such as; industry agriculture, medicine, etc. In this paper, we develop the electromagnetic optimization (EMO) algorithm based on levy function, EMO-levy, to enhance the EMO performance for determining the optimal multi-level thresholding of image segmentation. In general, EMO simulates the mechanism of attraction and repulsion between charges to develop the individuals of a population. EMO takes random samples from search space within the histogram of image, where, each sample represents each particle in EMO. The quality of each particle is assessed based on Otsu’s or Kapur objective function value. The solutions are updated using EMO operators until determine the optimal objective functions. Finally, this approach produces segmented images with optimal values for the threshold and a few number of iterations. The proposed technique is validated using different standard test images. Experimental results prove the effectiveness and superiority of the proposed algorithm for image segmentation compared with well-known optimization methods.
IS - Special Issue on Artificial Intelligence Applications
M1 - 4
N2 - Image segmentation is considered one of the most important tasks in image processing, which has several applications in different areas such as; industry agriculture, medicine, etc. In this paper, we develop the electromagnetic optimization (EMO) algorithm based on levy function, EMO-levy, to enhance the EMO performance for determining the optimal multi-level thresholding of image segmentation. In general, EMO simulates the mechanism of attraction and repulsion between charges to develop the individuals of a population. EMO takes random samples from search space within the histogram of image, where, each sample represents each particle in EMO. The quality of each particle is assessed based on Otsu’s or Kapur objective function value. The solutions are updated using EMO operators until determine the optimal objective functions. Finally, this approach produces segmented images with optimal values for the threshold and a few number of iterations. The proposed technique is validated using different standard test images. Experimental results prove the effectiveness and superiority of the proposed algorithm for image segmentation compared with well-known optimization methods.
PY - 2019
SP - 102
EP - 112
T2 - International Journal of Interactive Multimedia and Artificial Intelligence
TI - Multilevel Thresholding for Image Segmentation Using an Improved Electromagnetism Optimization Algorithm
UR - http://www.ijimai.org/journal/sites/default/files/files/2018/09/ijimai_5_4_12_pdf_62616.pdf
VL - 5
SN - 1989-1660
ER -