| 基于骨架的线状对流系统客观量化识别算法研究 |
| |
| 引用本文： | 盛杰,郑永光,沈新勇,张小雯.基于骨架的线状对流系统客观量化识别算法研究[J].大气科学,2020,44(6):1291-1304. |
| |
| 作者姓名： | 盛杰 郑永光 沈新勇 张小雯 |
| |
| 作者单位： | 1.南京信息工程大学气象灾害教育部重点实验室/气候与环境变化国际合作联合实验室/气象灾害预报预警与评估协同创新中心，南京 210044 |
| |
| 基金项目： | 国家重点研发计划项目2018YFC1506104、2018YFC1507506、2017YFC1502003，国家自然科学基金项目41790471、41530427、41975054、41930967，中国科学院战略性先导科技专项XDA20100304 |
| |
| 摘 要： | 本文将计算机图形学骨架概念应用到气象学领域，发展了回波图像预处理、骨架修剪处理以及长宽比量化处理技术，该方法能自动识别出雷达回波拼图中符合气象学标准的线状对流系统（quasi-linear convective systems，QLCSs）。首先结合2016年黄淮地区一次双QLCSs过程给出了基于骨架的QLCSs客观量化算法的具体技术流程，然后利用该方法对2016年6月安徽地区的QLCSs进行客观筛选，并进一步量化识别QLCSs的移动特征，结合灾害天气实况与主观识别进行对比评估，结果表明:结合气象学标准改造的骨架图像识别算法，较好保留了气象回波形状信息，在准确量化对流系统长短轴的基础上，实现QLCSs的有效识别。而获得的量化移动矢量等特征，一方面可应用于致灾QLCSs的分类研究，为开展长序列统计及致灾机理分析提供个例识别方法和量化特征，另一方面也为QLCSs的短临监测预警业务提供新的思路。
|
| 关 键 词： | 线状对流系统 图像识别 骨架 雷达回波 量化特征 |
| 收稿时间： | 2019-09-10 |
Research on Skeleton-Based Objective Quantization and Identification Algorithm for Quasi-linear Convective Systems |
| |
| Institution: | 1.Key Laboratory of Meteorological Disaster, Ministry of Education / Joint International Research Laboratory of Climate and Environment Change / Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing 2100442.National Meteorological Center, Beijing 1000813.Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519082 |
| |
| Abstract: | In this paper, we apply the skeleton concept from the field of computer graphics to meteorology and the development of echo image preprocessing, skeleton pruning, and length-width ratio quantization techniques. The quasi-linear convection systems (QLCSs) in radar echo mosaics that conform to meteorological standards can automatically be identified using this method. First, we introduce in detail the identification algorithm based on a double QLCSs process in the Huang–Huai area in 2016. Then, we use this method to objectively identify the QLCSs in the province of Anhui in June 2016 and quantify the moving characteristics of the QLCSs. We then compare weather disaster observations with their subjective identification. The results show that information about the shape of meteorological echoes is well preserved using the skeleton image identification algorithm and effective identification of QLCSs is realized base on the accurate quantification of the long and short axes of the convection system. Quantitative characteristics such as moving vectors can be applied to the classification of QLCS disasters to provide an identification algorithm and quantitative features for long–series statistics as well as an analysis mechanism for weather disasters. In addition, this concept provides a new technique for monitoring and warning by QLCSs. |
| |
| Keywords: | |
|
| 点击此处可从《大气科学》浏览原始摘要信息 |
| 点击此处可从《大气科学》下载免费的PDF全文 |
