Improved image segmentation method based on optimized higher-order polynomial

Image segmentation is an essential task in images analyzing images. At the pixel level, the images characterize as:1) an object from its background, or 2) the rest of the objects from each other. Therefore, image segmentation refers to applying a vast diffusion over several domains. This article proposes an improved image segmentation method. To get the segmented image, we determine the multi-thresholds by minimizing a higher-order polynomial curve fitting for an image and choosing the best threshold. The method can be utilized on grey images. The findings demonstrate that our method is effective and robust in image segmentation.