马尔科夫随机场MRF线性可变权重图像分割方法

Markov Random FieldLiner Variable Weight Image Segmentation Method

在对图像进行分割时,传统MRF模型没有考虑到像素间的相互关系,这样会使得分割不够准确。为此,本文提出了一种MRF线性可变权重图像分割方法。它在标记场和特征场中加入了邻域像素间的强度信息,从而可以有效运用图像空间信息。然后将指数型可变权重参数改为线性可变权重参数,来连接标记场和特征场,加快了分割结果更新速度,增大了势函数的选择范围。实验显示,当用改进算法分割不同类型的图像时,本文提出的算法在分割结果的准确性和区域一致性上,更具有效性和鲁棒性。不管是在分割速度上还是图像处理效率上,都有了很大的提升。

The traditional MRF model does not account for the interconnection between the pixels,which makes it less accurate in image segmentation. A MRF linear variable weight image segmentation method is proposed. First,it adds the intensity information among the neighboring pixels in the marked field and the characteristic field,so that the image spatial information can be effectively used. Then,the exponential variable weight parameter will be changed to the linear variable weight parameter to connect the marker field and the characteristic field,which can accelerate the update speed of the segmentation results and increase the selection range of the potential function.The experiment results show that the algorithm is more effective and robust in the accuracy and regional consistency of the segmentation results when segmenting different types of images. To sum up,there has been a great improvement both in the segmentation speed and image processing efficiency.