论文部分内容阅读
合成孔径雷达SAR(Synthetic Aperture Radar)图像的非平稳性是SAR海冰图像自动解释的主要障碍,入射角效应是导致海冰图像特征不稳定的主要因素之一。基于Radarsat-1 ScanSAR模式数据,本文提出了一种集成入射角效应校正步骤的分割算法。该方法综合考虑了入射角效应、斑点噪声等不确定因素,经由像素到区域再到大尺度区域这一途径,把区域聚类、类尺度上的入射角效应校正以及区域合并等操作组合起来,以有效提高分割算法对非平稳性的适应能力。针对巴芬湾和和波斯尼亚湾ScanSAR模式图像的实验表明,本文提出的方法可有效提高分割准确性。
Synthetic Aperture Radar image nonstationarity is the main obstacle for the automatic interpretation of SAR sea ice images. The incidence angle effect is one of the main factors that cause the instability of sea ice image features. Based on the Radarsat-1 ScanSAR model data, this paper presents a segmentation algorithm that integrates the correction steps of incident angle effects. The method takes into account the uncertainties such as incident angle effect, speckle noise and so on. Through the combination of pixel-to-area and large-scale area, this method combines regional clustering, correction of incident angle effect on class scale, In order to effectively improve the adaptability of segmentation algorithm to non-stationary. Experiments on ScanSAR model images of Baffin Bay and Bosnia have shown that the proposed method can effectively improve the segmentation accuracy.