Computer Engineering and Applications ›› 2010, Vol. 46 ›› Issue (21): 171-174.DOI: 10.3778/j.issn.1002-8331.2010.21.049

• 图形、图像、模式识别 • Previous Articles     Next Articles

Mesocyclone recognition based on new flooding algorithm

PAN Yun-hong1,ZHANG Yun1,WU Zhi-fang2,ZHONG Ling1   

  1. 1.Faculty of Automation,Guangdong University of Technology,Guangzhou 510006,China
    2.Guangzhou Central Observatory,Guangzhou 510080,China
  • Received:2009-08-24 Revised:2009-10-09 Online:2010-07-21 Published:2010-07-21
  • Contact: PAN Yun-hong

基于浸水模拟改进算法的中气旋自动识别

潘运红1,章 云1,伍志方2,钟 灵1   

  1. 1.广东工业大学 自动化学院,广州 510006
    2.广州中心气象台,广州 510080
  • 通讯作者: 潘运红

Abstract: A new recognition algorithm based on mathematical morphology is presented for the characteristics of mesocyclone in Doppler radar images.In order to effectively identify the extreme regions in a mesocyclone,a new flooding algorithm is proposed based on the regional attributes bound.The proposed algorithm well accords with the visual experience of professionals.Experiments demonstrate the effectiveness of the new algorithm.

Key words: flooding, extrema, mesocyclone recognition

摘要: 针对多普勒天气雷达图像中中气旋的特征,提出了一种基于数学形态学的自动识别算法。为了能有效识别出组成中气旋的一对极值区域,借鉴浸水模拟分水岭算法的思想,构造了一种结合区域属性约束的改进浸水模拟算法用于中气旋识别。新方法比较符合气象人员的视觉经验,充分利用中气旋的图像特征以满足特定的识别需求。对实测数据的识别结果表明了该方法的有效性。

关键词: 浸水模拟, 极值区域, 中气旋识别

CLC Number: