X射线环焊缝图像降噪及增强方法研究

为了降低X射线焊缝图像中的椒盐和高斯混合噪声,在结合中值和均值滤波优点的基础上,提出一种基于噪声强度概率识别滤波器,首先对两种噪声分离,然后根据像素点受噪声污染的概率大小,选择不同的滤波窗口,这样既保留了图像的细节又有针对性的除去了噪声;在增强部分,提出一种非线性图像增强方法,具有自动针对焊缝区域增强而非背景区域,可以有效增强环焊缝X射线图像至个人最佳视觉感受值,较传统方法增强效果好且更符合工业现场需求,最后,通过实验验证了所提降噪及增强方法的可行性及优势。

Research on theNoise reduction and enhancement of X-ray girth weld image

In order to reduce the mixed noise of salt and pepper and Gauss in X-ray weld image,based on the advantages of median filter and mean filter,a filter based on the probability of noise intensity recognition is proposed.Firstly,the two kinds of noise are separated,different filter windows are selected according to the probability of noise pollution of pixel points,which preserves the details of the image and removes the noise explicitly. In the enhancement part,this paper proposes a non-linear image enhancement method that can automatically enhance the girth weld X-ray image to the best personal visual perception value instead of the background area. The enhancement effect is better than the traditional method and more in line with the industrial field’s needs. Finally,the feasibility and advantages of the noise reduction and enhancement method proposed in this paper are verified by experiments.