摘要
为了提升基于马尔科夫随机场理论(MRF)的SAR海冰分割精度,文中定义SAR海冰图像中海冰和海水的均值差与均方差之差的比值为图像质量因子Q。通过改变SAR海冰样本图像中冰或水的灰度值,得出SAR海冰MRF分割精度与SAR海冰图像质量因子Q的关系曲线,用指数函数模型拟合并建立方程。定量化的描述了SAR海冰MRF分割效果与SAR海冰图像质量的关系,从而指导通过提升SAR海冰图像质量来提高SAR海冰识别效果的研究。
In order to improve SAR sea ice image segmentation precision of MRF,the SAR sea ice image quality factor Q was defined as the ratio between the mean difference and the mean square deviation difference of the sea ice and sea wa ter. Changed the gray value of the sea ice or sea water in SAR sea ice image,we got a curve between image segmentation precision of MRF and SAR sea ice image quality factor Q and then used exponential function fitting this curve and established an equation. Quantitatively described the relationship between SAR sea ice image segmentation ldrecision of MRF and image quality can guide deep research on improving SAR sea ice image segmentation precision of MRF by improving SAR sea ice image quality.
出处
《国外电子测量技术》
2013年第3期31-35,共5页
Foreign Electronic Measurement Technology
