针对X射线检测复合材料原始图像具有灰度区间窄、缺陷边缘模糊、图像噪声多、缺陷特征有时被淹没的特点.文中提出了一种基于软阈值的小波图像增强方法,图像经过小波分解后可以得到一系列不同尺度上的子带图像.在不同尺度的子带图像上进行基于软阈值滤波的细节系数增强.再进行图像重构.即可得到增强后的X射线检测图像。实验表明.该方法可以有效地增强图像的细节信息,抑制噪声,改善图像的视觉效果。
Focused on the problems of narrow gradation sector, fuzzy flaw edge,many image noise, the flaw characteristics were submerged sometimes for material X-ray inspection, a method of wavelet image enhancement based on soft threshold was advanced. The DWT decomposes an image into a set of different smaller scale images, then detail coefficients were enhanced based on the soft threshold filter, then the wavelet inverse transform can be obtained. Experimental results indicate the method can effectively enhance image details, silence noise and improve visual effect of image.