Context-Sensitive Thresholding Technique Using ABC for Aerial Images
Image anatomization is a remarkable notch in image processing which entails the scrutinization of the number of non-overlapping and homogeneous regions that exist in the input image. Thresholding is the most popular algorithm of image segmentation. In this article, the authors have utilized energy curve to incorporate spatial contextual information to inspect the regions where most favourable threshold(s) exist. The thresholding technique automatically computes the count of objects present in input image. To determine the optimal thresholds present in the image, artificial bee colony algorithm has been deployed. The results achieved have been compared with GA-based technique to ensure the efficacy of the proposed technique.
KeywordsArtificial bee colony (ABC) Image segmentation Optimization Thresholding
- 19.Goldberg, D., Holland, J.H.: Genetic Algorithms in Search, Optimization, and Machine Learning (1989)Google Scholar
- 20.Davis, D.L., Bouldin, D.W.: A cluster separation measure. IEEE Trans. Pattern Anal. Mach. Intell. PAMI-1(2), 224–227 (1979)Google Scholar