Multilevel Crop Image Segmentation Using Firefly Algorithm and Recursive Minimum Cross Entropy

Image segmentation plays an important role in computer vision technology and agriculture is one of their applications. The crop images present near the vicinity are complex and dense. Hence, multilevel thresholding of such crop images is a tedious task. In this paper, we propose multilevel thresholding of crop images using recursive minimum cross entropy and firefly algorithm. The firefly is based on the social behavior of the swarm of fireflies and bioluminescent information-sharing phenomena. It is a swarm-based algorithm, which offers a better search mechanism to find the optimum threshold value. The performance of the proposed method is observed over ten complex background crop images and compared with the wind-driven optimization algorithm. The better fidelity parameters evidence the superiority of the proposed method.