基于数字图像处理的海冰参数识别与统计分析

发布时间：2017-12-27 11:24

本文关键词：基于数字图像处理的海冰参数识别与统计分析　出处：《大连理工大学》2016年硕士论文　论文类型：学位论文

【摘要】：海冰是一种常见自然情况,在全球气候及海洋生态系统中均扮演着十分重要的角色。在海冰存在区域,极易出现浮冰,浮冰在风和海流作用下,具有较高速度,对极区、副极区和高纬度寒区海域的海上钻井平台、渔业养殖、船舶航行和人类活动带来严重的影响。然而,随着科技发展及极区蕴含丰富的自然资源,人类对于寒区利用和开发的活动呈现出上升的趋势,例如商用或科研考察航行和寒区海岸基础建设等,尤其是近海油田的开发,更是趋于常态。为了保障寒区作业的安全,从20世纪50年代开始针对海冰的物理力学性质、海冰漂移轨迹、海冰生消等方面进行了一系列研究。现行海冰监测主要包括：沿岸基础力学实验、油气作业区现场定点监测、无人机航测和破冰船监测。其中,基于海洋平台的现场定点监测具有误差小、范围广、连续性好、准确性高等优点,为冰期油气平台正常运转提供了及时可靠的现场海冰数据。我国自1990年开始对渤海油气田开发以来,在JZ20-2海洋平台上逐渐发展了健全了海冰现场监测系统,主要包括：海冰数字图像监测系统、海冰雷达监测系统、卫星遥感监测系统,对该海域的海冰速度、密集度、冰厚进行了系统、长期的现场监测,同时通过振动探头、应力盒、气象仪等设备完成对油气平台结构的冰激振动、冰荷载及该海域的气象、水文要素进行同步监测。为了更好的分析海冰在JZ20-2海域的漂移轨迹,本文在2012-2013年冬季,对JZ20-2海域的海冰漂移轨迹采用了全冰期雷达定点监测。通过海冰雷达图像处理系统进行处理,得到了该海域每隔10 min的海冰冰速与冰流向,共采集有效数据样本6883组。在此监测数据基础上,可对该海域冰速的概率分布进行统计分析,并确定不同天气条件下的海冰漂移轨迹,为合理的通过现场监测数据预测该海域海冰漂移行为提供了依据。本文通过对海冰卫星遥感图像研究,从中可以准确的提取出卫星遥感图像中的海冰浮冰位置、尺寸、密集度等信息,该系统采用去噪、腐蚀、图像二值化等算法,可以为渤海海冰的数值计算提供相应的海冰信息,通过数据库完成对海冰边界的提取工作。为了更好提取海冰参数,同时与雷达提取海冰数据对应,本文针对JZ20-2海域海冰现场图像监测的10年数据,得到该海域近十年的总冰期、有效冰期、各年的海冰情况,对JZ20-2平台附近海冰参数概率密度统计,得到该海域历年的海冰工况。同时,专门统计了2009-2010年海冰参数与平台振动之间关系,为该海域冬季采油平台安全作业提供相应的参考。除此之外,本文针对第32次南极科考的机会,在“雪龙”号安装了图像采集装置,对沿途的海冰信息进行记录。通过海冰图像密集度提取软件对沿途的海冰密集度信息进行提取,得到了沿途海冰密集度信息,为南极科考积累的大量信息。
[Abstract]:Sea ice is a common natural situation and plays a very important role in the global climate and the marine ecosystem. Ice floes are prone to ice floes in the areas where sea ice is present. The floating ice has a high speed under the action of wind and ocean currents, which has a serious impact on the offshore drilling platform, fishery culture, ship navigation and human activities in polar regions, sub polar regions and high latitude cold regions. However, with the development of science and technology and the rich natural resources in the polar region, human activities show an upward trend in the cold area utilization and development activities, such as commercial or scientific research, navigation and cold area coastal infrastructure construction, especially offshore oil field development. In order to ensure the safety of cold area operations, a series of researches were carried out in 1950s on the physical and mechanical properties of sea ice, the drift of sea ice and the generation and elimination of sea ice. The current sea ice monitoring mainly includes the coastal basic mechanics experiment, the site fixed-point monitoring in the oil and gas operation area, the unmanned aerial vehicle aerial survey and the ice breaking ship monitoring. The site based fixed point monitoring based on offshore platform has the advantages of small error, wide range, good continuity and high accuracy. It provides timely and reliable field sea ice data for the normal operation of ice platform. In China since the beginning of 1990 since the Bohai oil and gas field development, in the JZ20-2 platform on the gradual development of a sound monitoring system, mainly including: sea ice sea ice digital image monitoring system, sea ice radar monitoring system, satellite remote sensing monitoring system for the sea ice velocity, concentration, ice thickness, on-site monitoring system for a long time, at the same time through vibration probe, stress box, meteorological instrument of oil and gas platform structure of ice induced vibration, ice load and the waters of the meteorological and hydrological synchronous monitoring. In order to better analyze the drift track of sea ice in the JZ20-2 sea area, in the winter of 2012-2013 years, the sea ice drift track in JZ20-2 sea area was monitored by full glacial radar. Through the sea ice radar image processing system, the ice speed and ice flow direction of sea ice every 10 min in the sea area were obtained, and 6883 groups of effective data samples were collected. Based on this monitoring data, the probability distribution of ice speed in this area can be statistically analyzed, and the sea ice drift trajectory under different weather conditions can be determined, which provides a basis for predicting the sea ice drift behavior in the sea area reasonably through field monitoring data. Based on the study of satellite remote sensing image can be extracted from sea ice, sea ice floe position, remote sensing image size, intensity of satellite information accurately, the system adopts denoising, image binarization and corrosion algorithm, can provide the corresponding information for the Bohai sea ice numerical calculation, through the database to complete extraction the sea ice boundary. In order to extract the parameters of sea ice, and sea ice extraction data corresponding with radar, aiming at image monitoring of JZ20-2 sea ice field data of 10 years, and the total glacial waters for nearly ten years, the annual effective glacial ice conditions, on the JZ20-2 platform near the sea ice parameters of probability density statistics, the waters over the sea ice conditions. At the same time, the relationship between the sea ice parameters and the vibration of the platform for 2009-2010 years is statistically analyzed to provide a reference for the safety operation of the winter oil extraction platform in this sea area. In addition, this article, in view of the opportunity of the thirty-second Antarctic expedition, installed the image collection device on the "Snow Dragon", and recorded the sea ice information along the way. The sea ice concentration information along the route was extracted through the sea ice image density extraction software, and the sea ice concentration information along the way was obtained, which accumulated a lot of information for the Antarctic expedition.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P731.15;P714.1

【参考文献】