系留缆绳损伤检测图像去噪方法研究

Research of Image De-Noising Method in Mooring Line's Defect Detection

为了提高气球载雷达系留缆绳损伤检测中的图像质量,提出了一种多尺度多结构形态学的小波融合去噪方法。首先对噪声图像进行小尺度多结构元素的形态开-闭(OC)运算,然后再进行较大尺度多结构元素的形态开-闭运算,利用小波融合技术将运算结果进行融合,最后进行了仿真验证和性能分析。结果表明,采用该方法得到的峰值信噪比(PSNR)普遍高于传统的去噪方法,在抑制噪声的同时较好地保护了图像的细节和边缘信息,尤其在噪声密度较大时,优势更加明显,峰值信噪比能高出十多dB,而且对随机噪声和椒盐噪声均能获得良好效果,具有较强的适用性。

Aimed at improving images＇ quality in defect detection of balloon-borne radar＇s mooring line, a de-noising method that combines multi-scale and multi structure morphology with wavelet fusion is proposed. Firstly, the noising image is pro- cessed by morphologic open closing operator using smaller multi-structure elements and bigger multi-structure elements. Then fusing the two results by wavelet fusion technique, last simulating the method and analyzing its performance. Simula- tion results show that this method is higher than those traditional methods in PSNR （Peak Signal to Noise Ratio）, the meth- od can both suppress noises and protect details and edges effectively. Especially when noise density is higher, the advantage of the method is more obvious, and the PSNR could be higher than those methods to more than ten decibels. At the same time this method can be adapted to both random noise and impulse noise.