Infrared object segmentation method based on weighted information entropy and markov random field

The invention relates to an infrared object segmentation method based on weighted information entropy and markov random field. The infrared object segmentation method is technically characterized by comprising the following steps of: firstly, utilizing the weighted information entropy to obtain an interested region, afterwards utilizing a gauss markov random field to build a model in an interested region to seize contextual information of images, and adopting an ICM (Interdisciplinary Contest in Modeling) optimization algorithm in iteration to greatly reduce the solution space and the noise interference, thereby reducing iterative steps, obviously improving the segmentation precision and greatly enhancing the segmentation speed. The validity of the invention is validated by a plurality ofinfrared object segmentation experiments.