Handling and quantifying two types of uncertainties, spatial and inherent, in land-cover segmentation using multiple satellite images constitutes a primary concern within the domain of segmenting multiple remote sensing images. Despite the comprehensive examination of the spatial uncertainty, the lack of a mathematical model addressing the inherent uncertainty in satellite images is notable. This paper endeavors to address this gap by considering and focuses on the latter. Leveraging multiple high-resolution panchromatic remote sensing (HR-PRS) images as input data, the study aims to model all procedures in the image formation and preprocessing stages (applied to each input image before segmentation) as the producing factors of inherent uncertainty in the segmentation process. The input images, conceptualized as events in random processes, undergo fusion using the mean estimator, referred to as the proposed "ultrafusion" method in this paper. The study demonstrates that the results of ultra-fusion can be effectively modeled by normal-type fuzzy numbers. The parameters of these fuzzy numbers are estimated using the maximum likelihood method. Subsequently, Fuzzy C-Means (FCM), as the most renowned clustering method, is employed for segmentation. Analytical comparisons reveal the notable performance of the proposed algorithm in terms of accurate segmentation and computational efficiency. Affirming the efficacy of the proposed approach, simulation results validate the advantages by improving the Overall Accuracy, Kappa, and F1-Score indices by about 0.86, 0.52, and 1.03 percent respectively, in compared to the most recent and similar method.