摘要
基于国产单组份压力敏感涂料预先校准试验,以自主编制的涂料校准与特性分析图像后处理软件为主要分析工具,对PSP图像平均、滤波以及无效采集图像剔除等图像后处理方法进行了较为深入地研究,研究分析结果表明:一些随机因素或操作失误会造成某些PSP采集图像为跳点,图像后处理中,应将无效采集图像剔除;与高斯滤波、平均滤波相比,中值滤波可以更好地抑制图像的椒盐噪声,同时,滤波窗口大小对图像滤波效果有较大影响;图像平均可以有效地抑制CCD相机的随机噪声,同时,存在着最佳的图像平均幅数,超过最佳图像平均幅数,随机噪声的抑制效果变化不再明显,反而会降低试验效率。
Herein,prior calibration experiments using domestic single-component fluorescent pressure sensitive paint（PSP）has been conducted to research the PSP image post-processing methods,including image average,filter and invalid images picking out technique,based on PSP prior calibration system developed by Northwestern Polytechnical University and an image post-processing software（Paint Calibration and Performances Analysis Software,PCPAS）programmed by CARDC independently.The research results indicated that random factors or operation misplay may produce some invalid PSP images which should be picked out during the image post-processing.Moreover,comparing with the Gauss filter and average filter,median filter method can effectively eliminate the image salt and pepper noises.At the same time,the size of filter window will influence the filter effect to some extent.At last,it also can be concluded that image averaging can eliminate the image random noises effectively.Meanwhile,there is an optimum average frame number exited,the filter effect will almost have no more change when the image frames exceed the optimum average frame number.
出处
《空气动力学学报》
CSCD
北大核心
2014年第2期197-202,共6页
Acta Aerodynamica Sinica
基金
国家自然科学基金(11172316)
