Résumé: Satellites image segmentation is useful in a variety of fields. Professionals in related industries can gain a lot of important information by segmenting satellite images, but processing them is difficult due to the complex background and various locations of interest. The simplest basic picture segmentation approach is multilevel threshold segmentation, but the results can be unsatisfactory due to its computational complexity and high computation time. This paper presents an improved bald eagle search algorithm based on the standard bald eagle search algorithm. This method is used to find the best threshold values for the Renyi entropy multilevel threshold methods. To expand the diversity of the bald eagle population and the bald eagle's prey space, the strategy uses Chaotic Tent and Levy Fight method. The method is utilized to do multilevel segmentation of color satellite images, and the results suggest that it is capable of doing so. At the same time, this method has fewer iterations and can achieve better threshold values than typical methods such as standard bald eagle search optimization algorithm.

Langue: Anglais