An image segmentation algorithm based on Markov random field

Abstract

The invention discloses an image segmentation algorithm based on a Markov random field, belonging to the technical field of image segmentation. The invention studies a Markov image segmentation algorithm realized by a conditional iterative algorithm (ICM). The Markov image segmentation algorithm adopts random pre-classification to perform preliminary segmentation on the image. The invention has the disadvantage that the number of iterations is large, partial regions are easily trapped into a local optimal solution, and the segmentation accuracy of details such as edge contour is insufficient at the same time. But its theoretical basis is simple, the model is easy to implement, the coupling coefficient, self-defined classification number and iteration number can be changed, in order to achieve a more ideal segmentation results, the adjustment of different scene parameters is more flexible, having a better plasticity and use value. This method combines Markov random field image segmentation algorithm with graph-based image segmentation (GBIS) algorithm, which can overcome the shortcomings of the original segmentation algorithm to a certain extent, and obtain a better segmentation result.

Description

technical field [0001] The invention belongs to the technical field of image segmentation, and in particular relates to an image segmentation algorithm based on a Markov random field. Background technique [0002] Image segmentation technology has attracted much attention for many years, and has a wide range of applications in many fields such as image recognition, remote sensing, and medical diagnosis. As the basis and premise of image processing and analysis, the purpose of segmentation is to divide the image into several regions with specific properties. The segmentation results have a great influence on the follow-up work, and good results will provide great help to the follow-up work. It goes well, but bad results may make it impossible to carry out follow-up work, or even analyze the specific reasons for the failure of the experiment. [0003] Many methods have been proposed to solve this problem. For example, feature-based methods such as clustering and thresholding...

AI Technical Summary

Problems solved by technology

[0005] At the same time, MRF also has obvious disadvantages, such as easy to fall into a local optimal solution, and the model parameters are not suitable for accurate estimation. At the same time, the small number of model parameters also makes the solution speed pay a certain price.

Images