粒子图像测速中的散斑提取方法  

The Study on Speckle Extraction Method of Particle Image Velocimetry

作　　者：杜海[1] 李木国[1] 孟娟[2] 马乐[1] 

机构地区：[1]大连理工大学海岸和近海工程国家重点实验室,辽宁大连116024 [2]大连海洋大学信息工程学院,辽宁大连116024

出　　处：《实验室研究与探索》2015年第11期4-7,共4页Research and Exploration In Laboratory

基　　金：国家自然科学基金项目(编号61202253)

摘　　要：研究示踪粒子的成像特点,提出一种适用于粒子图像的散斑提取方法:根据目标连通域的统计特征进行自适应二值化处理;从多个径向判断目标区域是否存在混叠并辩识粒子;提取粒子的中心位置。该方法有效地将粒子图像特征融入到二值化阈值提取与粒子辨识中,从而正确提取粒子的中心。使用粒子图像进行了算法验证。结果表明,所提算法可以有效地对粒子图像中的散斑进行提取,拓展了粒子示踪法在应用时的空间深度。When one uses particle tracking method to do fluid researches within 3D space,the particle aliasing often causes tracking errors. In order to solve identification problem caused by particle mixture images,a new speckle extraction method is presented by studying on the distribution features of particle images. In the proposed algorithm,a new adaptive binarization is used to eliminate noise firstly. And,the mixture particles are distinguished by radius directions fitting method,and then particle centers are found. Using the proposed method,particles within a thickness laser slice can be detected correctly. At last,particle images are tested. The experimental results show that the proposed method is an effective and robust algorithm for extracting speckles in particle image.

关 键 词：粒子图像测速 散斑提取 二值化 混叠检测 

分 类 号：O353.5[理学—流体力学]