| 基于面向对象和模糊逻辑的SAR溢油检测算法 |
| |
| 引用本文： | 苏腾飞,李永香,李洪玉.基于面向对象和模糊逻辑的SAR溢油检测算法[J].海洋学报,2016,38(1):69-81. |
| |
| 作者姓名： | 苏腾飞 李永香 李洪玉 |
| |
| 作者单位： | 1.内蒙古农业大学 水利与土木建筑工程学院, 内蒙古自治区 呼和浩特 010018;国家海洋局第一海洋研究所, 山东 青岛 266061 |
| |
| 基金项目： | 国家自然科学基金(60890075)。 |
| |
| 摘 要： | 星载合成孔径雷达(Synthetic Aperture Radar,SAR)具有全天时、全天候的工作能力,已被众多学者认为是非常适合探测海面溢油污染的遥感器。然而在SAR影像中经常出现"类油膜"现象,这严重干扰了SAR溢油检测的精度。因此,如何有效区分SAR影像中的油膜和类油膜,对提升溢油检测精度具有重要意义。本文利用面向对象图像分析的方法,从20景ENVISAT ASAR影像中提取了较多的溢油和类油膜样本,对其基于对象的形状、物理和纹理特征进行了综合分析,找出了适合区分溢油和类油膜的特征量。利用特征分析的结论,本文建立了一种基于模糊逻辑的溢油检测算法。该算法可以有效区分SAR影像中的溢油和类油膜,还可以给出暗斑被判定为溢油的概率。溢油检测实验说明,本文方法能够得到令人满意的效果。
|
| 关 键 词： | 合成孔径雷达 特征分析 溢油分类 面向对象图像分析 模糊逻辑 |
| 收稿时间： | 1/6/2015 12:00:00 AM |
| 修稿时间： | 2015/4/27 0:00:00 |
Sea oil spill detection method by SAR imagery using object-based image analysis and fuzzy logic |
| |
| Su Tengfei,Li Yongxiang and Li Hongyu.Sea oil spill detection method by SAR imagery using object-based image analysis and fuzzy logic[J].Acta Oceanologica Sinica (in Chinese),2016,38(1):69-81. |
| |
| Authors: | Su Tengfei Li Yongxiang and Li Hongyu |
| |
| Affiliation: | 1.Inner Mongolian Agricultural University, College of Water Conservancy and Civil Engineering, Hohhot 010018, China;The First Institute of Oceanography, State Oceanic Administration, Qingdao 266061, China2.College of Geographical Science, Inner Mongolia Normal University, Hohhot 010022, China3.Inner Mongolian Agricultural University, College of Water Conservancy and Civil Engineering, Hohhot 010018, China |
| |
| Abstract: | Synthetic aperture radar (SAR),a sensor with all weather and day and night working capacity,has been widely considered as a powerful tool for sea surface oil spill detection. However,lookalikes frequently appear in SAR images,limiting the performance of SAR to detect oil spilled at sea. Thus it is important to study how to effectively differentiate oil spill from lookalike. By using Object-based Image Analysis (OBIA),a number of oil spill and lookalike samples are extracted from 20 scenes of ENVISAT ASAR images. The object-based geometric,physical and textural features of the samples are analyzed with the objective of determining the best feature variables for oil spill and lookalike separation. The conclusions derived from feature analysis are utilized for the construction of an FL-based oil spill classifier. The proposed method can effectively single out oil spill from lookalike,giving the crisp probability of a dark segment being oil spill at the same time. Oil spill detection experiment indicates that our method can produce satisfactory result. |
| |
| Keywords: | SAR feature analysis oil spill classification OBIA fuzzy logic |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《海洋学报》浏览原始摘要信息 |
|
点击此处可从《海洋学报》下载全文 |
|
