摘要
在可见光图像生成红外图像的过程中,图像分割至关重要。马尔可夫随机场(MRF)具有局部特性,由此特性建立了纹理特征的MRF模型。利用纹理的MRF模型,将参数的期望最大化用于该模型中的参数估计。最后将图像中的所有像素经该模型计算后得到纹理信息并分割图像。通过实验取得了较好的效果。
Image segmentation is the most important step in creating infrared image from optical image. Markov random field (MRF) has local characteristics. The MRF model of texture is based on it. The MRF model of texture is used, estimating the parameters in the model with expectation-maximization (EM) method, and all pixels in the image with the model are calculated, information about image texture, and segment the image are gotten. Satisfied result is can gotten by experimentation with this algorithm.
出处
《科学技术与工程》
2006年第6期768-770,共3页
Science Technology and Engineering
关键词
图像分割
MRF模型
纹理分析
Image Segmentation
MRF model
Texture analysis