遥感图像中尺度海洋锋及涡旋提取方法研究进展
发布时间：2018-11-20 07:33
【摘要】：目的中尺度海洋锋及涡旋均是重要的中尺度海洋环境特征。中尺度海洋锋及涡旋的提取及其时空分布、变化的研究对海洋生态系统的研究、渔业资源评估、渔情预报及军事等都有重要意义。遥感技术能在同一时间获取大面积海洋要素观测数据,遥感数据具有优良的连续性、同步性,因此遥感数据被广泛应用于中尺度海洋锋及涡旋提取的研究中。方法对基于遥感数据进行中尺度海洋锋提取的梯度法、Canny算法、小波分析法和基于引力模型的方法,以及涡旋的提取的Okubo-Weiss法(OW法)、Winding-Angle法(WA法)、基于海面高度的无阈值等值线法和Hybird Detection(HD法)进行总结和分析,并提出对中尺度海洋锋面及涡旋提取方法的见解及新思路。结果利用2014年2月南海北部海表温度(SST)数据,分别采用梯度法中的Gradient法、Sobel算法以及Canny算法对南海北部温度锋进行提取并得到该区域温度锋分布图。结果表明在多种锋面提取方法中,Canny算法具有较高的效率且其提取结果的连续性和精度更好。中尺度涡的提取方法中,WA法的提取结果具有更好的准确性。早期的中尺度涡提取方法忽略了多中心结构涡旋存在的情况,而后来的HD法能较好地识别多中心结构涡旋。结论阈值选取是中尺度海洋锋及涡旋提取的难点和提取结果好坏的关键。然而海洋要素图像弱边缘的特点使得传统边缘检测方法不一定适用于中尺度锋提取。文章通过对不同锋面及涡旋提取方法的总结与分析,为海洋锋面及涡旋提取的研究提供了参考依据。
[Abstract]:Objective Mesoscale ocean front and vortex are important characteristics of mesoscale marine environment. The extraction of mesoscale ocean fronts and vortices and their temporal and spatial distribution and variation are of great significance to the study of marine ecosystem, fishery resource assessment, fishery situation prediction and military affairs. Remote sensing technology can obtain large area ocean element observation data at the same time. Remote sensing data have excellent continuity and synchronism, so remote sensing data are widely used in the research of mesoscale ocean front and vortex extraction. Methods the gradient method, Canny algorithm, wavelet analysis method and gravity model based method were used to extract mesoscale ocean fronts from remote sensing data. Okubo-Weiss method (OW method) and Winding-Angle method (WA method) were used to extract vortex. Based on the sea surface height without threshold contour method and Hybird Detection (HD method, this paper summarizes and analyzes the method of mesoscale ocean front and vortex extraction, and puts forward some new ideas for mesoscale ocean front and vortex extraction method. Results based on the (SST) data of sea surface temperature in the northern South China Sea in February 2014, the temperature front of the northern South China Sea was extracted by the Gradient method, the Sobel algorithm and the Canny algorithm, respectively, and the temperature front distribution map was obtained. The results show that the Canny algorithm is more efficient and has better continuity and accuracy in many frontal extraction methods. In the mesoscale vortex extraction method, the WA method has better accuracy. The early mesoscale vortex extraction method ignored the existence of polycentric vortex, while the later HD method could better identify the polycentric vortex. Conclusion the threshold selection is the key to the extraction of mesoscale ocean fronts and vortices. However, because of the weak edge of ocean image, the traditional edge detection method is not suitable for mesoscale front detection. Through the summary and analysis of different frontal and vortex extraction methods, this paper provides a reference for the study of ocean frontal and vortex extraction.
【作者单位】： 中国水产科学研究院渔业资源遥感信息技术重点实验室;上海海洋大学海洋科学学院;
【基金】：国家自然科学基金项目(31602206) 上海市自然科学基金项目(16ZR1444700) 中央公益性科研院所基本科研业务费(2016T05) 上海科创项目(14DZ1205000)~~
【分类号】：TP751
本文编号：2344263
[Abstract]:Objective Mesoscale ocean front and vortex are important characteristics of mesoscale marine environment. The extraction of mesoscale ocean fronts and vortices and their temporal and spatial distribution and variation are of great significance to the study of marine ecosystem, fishery resource assessment, fishery situation prediction and military affairs. Remote sensing technology can obtain large area ocean element observation data at the same time. Remote sensing data have excellent continuity and synchronism, so remote sensing data are widely used in the research of mesoscale ocean front and vortex extraction. Methods the gradient method, Canny algorithm, wavelet analysis method and gravity model based method were used to extract mesoscale ocean fronts from remote sensing data. Okubo-Weiss method (OW method) and Winding-Angle method (WA method) were used to extract vortex. Based on the sea surface height without threshold contour method and Hybird Detection (HD method, this paper summarizes and analyzes the method of mesoscale ocean front and vortex extraction, and puts forward some new ideas for mesoscale ocean front and vortex extraction method. Results based on the (SST) data of sea surface temperature in the northern South China Sea in February 2014, the temperature front of the northern South China Sea was extracted by the Gradient method, the Sobel algorithm and the Canny algorithm, respectively, and the temperature front distribution map was obtained. The results show that the Canny algorithm is more efficient and has better continuity and accuracy in many frontal extraction methods. In the mesoscale vortex extraction method, the WA method has better accuracy. The early mesoscale vortex extraction method ignored the existence of polycentric vortex, while the later HD method could better identify the polycentric vortex. Conclusion the threshold selection is the key to the extraction of mesoscale ocean fronts and vortices. However, because of the weak edge of ocean image, the traditional edge detection method is not suitable for mesoscale front detection. Through the summary and analysis of different frontal and vortex extraction methods, this paper provides a reference for the study of ocean frontal and vortex extraction.
【作者单位】： 中国水产科学研究院渔业资源遥感信息技术重点实验室;上海海洋大学海洋科学学院;
【基金】：国家自然科学基金项目(31602206) 上海市自然科学基金项目(16ZR1444700) 中央公益性科研院所基本科研业务费(2016T05) 上海科创项目(14DZ1205000)~~
【分类号】：TP751
【相似文献】
相关期刊论文 前2条
1 陈标,张本涛,何伟平;利用星载SAR图像检测海洋锋的方法[J];遥感技术与应用;2002年04期
2 张伟;曹洋;罗玉;;一种基于Canny和数学形态学的海洋锋检测方法[J];海洋通报;2014年02期
，本文编号：2344263
本文链接：https://www.wllwen.com/guanlilunwen/gongchengguanli/2344263.html
