采用小波变换方法对复合材料红外热波检测的图像增强进行研究。先对小波变换图像增强的原理进行介绍,将其分为初始化、分解、非线性增强和重构等4个步骤。然后对检测图像进行3层小波分解。随后,分别对3层分解系数进行增强处理,并通过阀值分割的方法对比增强效果,结果表明,对第2层分解系数进行增强处理的效果最佳。最后将最终增强处理图像与原图进行对比,证明了该方法的有效性,为复合材料的热波检测图像增强找到一种有效方法。

In order to resolve the problem of image enhancement of thermal wave NDT of composite materials, wavelet transform was introduced. Firstly, the target of image enhancement was cleared and definite and then, the frequency bonds distribution of image information were analyzed, and it was indicated that defect information was distributed in low-frequency and high-frequency bonds. On this base, the enhancement program was designed, and it was divided into 4 steps， initialization，decomposition, nonlinear enhancing and reconfiguration. After that, 3 layers wavelet decomposition was done to testing image, and 4 characters of decomposed coefficients were concluded. And then, these coefficients were done enhancement processing, and enhancement effects of them were compared by means of value segmentation method. It was witnessed that, the enhancement effect of the second decomposition coefficients was best. Finally, the best enhancement image was compared with original figure, and the effectiveness of the method was proved. An effective image enhancement method was found for thermal wave NDT of composite materials.