刘祥, 熊健, 马护生, 周强, 陈柳生, 王红彪, 黄辉, 陈植. 温敏漆校准及图像后处理方法研究[J]. 实验流体力学, 2020, 34(4): 53-61. DOI: 10.11729/syltlx20190054
引用本文: 刘祥, 熊健, 马护生, 周强, 陈柳生, 王红彪, 黄辉, 陈植. 温敏漆校准及图像后处理方法研究[J]. 实验流体力学, 2020, 34(4): 53-61. DOI: 10.11729/syltlx20190054
LIU Xiang, XIONG Jian, MA Husheng, ZHOU Qiang, CHEN Liusheng, WANG Hongbiao, HUANG Hui, CHEN Zhi. The calibration and image post-processing method research of temperature-sensitive paint[J]. Journal of Experiments in Fluid Mechanics, 2020, 34(4): 53-61. DOI: 10.11729/syltlx20190054
Citation: LIU Xiang, XIONG Jian, MA Husheng, ZHOU Qiang, CHEN Liusheng, WANG Hongbiao, HUANG Hui, CHEN Zhi. The calibration and image post-processing method research of temperature-sensitive paint[J]. Journal of Experiments in Fluid Mechanics, 2020, 34(4): 53-61. DOI: 10.11729/syltlx20190054

温敏漆校准及图像后处理方法研究

The calibration and image post-processing method research of temperature-sensitive paint

  • 摘要: 涂料校准误差是温敏漆实验技术中的重要误差源。针对国产615/EP配方温敏漆实施校准实验,系统研究了校准过程中涉及的CCD曝光时间,图像平均幅数,噪点的强度、体积和数量,图像清晰度和模数转换位数等技术参数对校准精准度的影响。研究结果表明:噪点的强度、体积和数量对涂料校准精准度影响很大。在噪点数量少、强度弱的取图区域中,图像清晰度对校准精准度影响很小,单幅图像即可保证校准精准度的要求。为了获得较高的校准精准度,对于滨松ORCA-R2型CCD相机,建议采集图像灰度值大于相机满井值的32%,同时要求温敏漆风洞实验与校准实验中采集图像的位数保持一致。

     

    Abstract: Temperature-sensitive Paint (TSP) calibration error is an important error source in the TSP experimental technique. Based on the domestic 615/EP TSP, the effects of some calibration technical details including the CCD exposure time, image average quantity, noise level, image definition and analog-to-digital (A/D) conversion digit on the calibration precision are studied. The research results show that, the image noise level has a significant effect on the calibration precision. In an image region of low noise level, the image definition and the image average quantity have less significant influence on the calibration precision. In order to obtain high-precision calibration results, the image gray should exceed 32% of the CCD full-well capacity for the Hamamatsu CCD camera, and the tunnel test and calibration test should have the same number of digits for the image A/D conversion.

     

/

返回文章
返回