| 基于Hough变换原理的海冰厚度识别方法 |
| |
| 引用本文： | 张培宣,陈晓东,孔帅,季少鹏,季顺迎.基于Hough变换原理的海冰厚度识别方法[J].海洋学报,2022,44(7):161-169. |
| |
| 作者姓名： | 张培宣 陈晓东 孔帅 季少鹏 季顺迎 |
| |
| 作者单位： | 1.大连理工大学 工业装备结构分析国家重点实验室，辽宁 大连 116023 |
| |
| 基金项目： | 国家重点研发计划重点专项(2018YFA0605902)；国家自然科学基金（42176241，52101300，12102083）；中央高校基本科研业务费（DUT21LK03）；水动力学重点实验室稳定支持基金；国家级大学生创新创业训练计划支持项目（20211014110061）。 |
| |
| 摘 要： | 作为主要海冰参数之一的海冰厚度对海冰灾害评估和极地船舶与冰区海洋工程结构设计具有重要意义。采用船侧视频图像对海冰厚度进行自动识别是提取海冰参数的重要方式。本文采用基于Hough变换的机器视觉方法对海冰翻转过程中的表面轮廓线进行识别，从而自动获取海冰厚度参数。根据海冰图像特征制定了图像边缘识别?近似线段识别?海冰轮廓线段组识别的计算流程。在线段组识别过程中，根据海冰的几何特征建立了由夹角、长度及间距参数相关联的3个识别参数所组成的判断条件。为验证方法的可靠性，将该方法用于“雪龙”号第八次北极科考的走航实测数据中，结果表明，3个识别参数均具有最优阈值。当低于最优值时提高阈值可增加有效识别率；而高于最优值时提高阈值则会导致误判率增大，采用最优阈值可使冰厚识别率达到90%以上。因此，采用基于Hough变换的冰厚识别方法可实现对海冰厚度的实时监测。
|
| 关 键 词： | 冰厚识别 海冰图像 Hough变换 图像轮廓线识别 |
| 收稿时间： | 2021-09-29 |
Research on sea ice thickness identification method based on Hough transform principle |
| |
| Institution: | 1.State Key Laboratory of Industrial Equipment Structure Analysis, Dalian University of Technology, Dalian 116023, China2.DUT-BSU Joint Institute, Dalian University of Technology, Dalian 116023, China3.China Ship Scientific Research Center, Wuxi 214082, China |
| |
| Abstract: | Sea ice thickness is one of the main sea ice parameters. Automatic recognition of sea ice thickness in video is a significant component of sea ice parameters extraction. In this paper, the machine vision method based on Hough transform is used to recognize the surface contour of sea ice, so as to obtain the sea ice thickness parameters. According to the characteristics of sea ice image, the overall recognition process is divide into image edge recognition, approximate line segment recognition and sea ice contour segment group recognition. In the process of line segment identification, three parameters of line segment group including angle, length and spacing are established based on the geometric characteristics of sea ice. In order to verify the reliability of the method, this method is applied to analysis the field survey data of Xuelong icebreaker’s eighth Arctic expedition. The results show that the three parameters have the optimal threshold value. When it is lower than this value, increasing the threshold will increase the effective recognition rate; when it is higher than this value, increasing the threshold will increase the false recognition rate. The ice thickness recognition rate can reach more than 90% by using the optimal threshold. Therefore, the ice thickness identification method based on Hough transform can realize the real-time monitoring of sea ice thickness. |
| |
| Keywords: | |
|
| 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》下载免费的PDF全文 |
|
