EAST装置高速VUVI系统对ELM丝状结构的边缘检测

Edge detection of ELM filament structures by high-speed VUVI system on EAST

EAST装置高速真空紫外成像(VUVI)诊断系统是边界局域模(ELM)二维丝状结构研究中的一项重要诊断设备。受工作环境以及高速采集的影响,VUVI系统丝状结构原始图像数据的信噪比相对较差。采用块匹配三维滤波(BM3D)算法、背景差分和数学形态学处理的方法对原始数据进行处理。使用BM3D算法去除所采集图像的噪声,增加信噪比;使用背景差分法提取丝状结构;采取数学形态学处理图像并获得丝状结构的边缘位置和倾斜角度。模拟与实验数据分析表明,该方法能有效提取VUVI诊断获得的丝状结构边缘细节,获得丝状结构倾斜角等信息,并提高了实验数据的利用率。

High-speed vacuum ultraviolet imaging(VUVI)system is an important diagnostic facility in the study of the two-dimensional filament structures of ELMs on EAST.Affected by the working environment and high-speed acquisition,the signal-noise ratio of the original image data of the filament structures from the VUVI system is relatively poor.In this paper,the original data are processed using block-matching and 3D filtering(BM3D)algorithm,background subtraction,and morphology processing.BM3D algorithm is used to remove the noise and increase the signal-noise ratio of the collected images.The filament structures are extracted by the background subtraction.The edge position and pitch angle of the filament structures are obtained through image processing by morphology.Simulation and experimental data analysis show that the proposed method can effectively extract the details of filament structures edge obtained by VUVI,obtain the information such as the pitch angle of filament structures,and improve the utilization rate of experimental data.