论文部分内容阅读
针对合成孔径雷达海冰图像分割特征上存在的鲁棒性问题,提出一种新的马尔科夫随机场(MRF)分割方法.传统的用于分割的双分量MRF模型,需要训练数据来估计所需的模型参数,因此不适合于非监督分割.通过在两个分量间引入基于函数的权重参数,每个分量对整个系统的贡献大小赋予权重参数,使用模拟退火方案时考虑空间关系信息,从而改进参数.这种基于区域的MRF模型方法能够自动估计模型参数并产生精确的非监督分割结果.实验表明,本文方法比RK-means和C-MLL分割方法具有更准确的边缘定位能力和更快的分割速度.
Aiming at the robustness of synthetic aperture radar (ICAR) sea ice image segmentation, a new MRF segmentation method is proposed. The traditional two-component MRF model for segmentation requires training data to estimate Which is not suitable for unsupervised segmentation. By introducing a function-based weight parameter between the two components, the weight contribution of each component to the entire system is given to the weight parameter, and the spatial relationship information is taken into account when using the simulated annealing scheme Improve the parameters.This method based on region-based MRF model can automatically estimate the model parameters and produce accurate unsupervised segmentation results.Experiments show that the proposed method has more accurate edge localization ability and faster than RK-means and C-MLL segmentation methods Split speed.