中文摘要：

分析了红外热波检测图像中噪声生成机理,针对图像的噪声去除增强问题,提出了一种改进的非线性偏微分方程的热波检测图像去噪增强方法。对偏微分方程在图像处理应用中的理论基础进行研究,针对现有的偏微分方程模型在图像去噪中存在的问题,并结合热波图像的特点,改进相应的偏微分方程数学模型,在此基础上对获得原始热图进行编程处理,试验结果表明：经过处理后的图像信噪比提高,对比度得到了改善,为后续的缺陷提取及损伤识别奠定了基础。

英文摘要：

The noise mechanism of infrared thermal wave non-destructive testing is analyzed. In order to suppress the noise and enhance the image, a new image denoising algorithm for infrared thermal wave NDT based on the improved nonlinear partial differential equation （PDE） is proposed. Firstly, the theoretical basis of PDE for image processing was studied. To solve the existing problems of the PDE in image processing, and combined the characteristics of the ther- mal wave images, the mathematical model of PDE was improved, and then the original infrared image was processed by programming. The results show that the SNR of image and the contrast are improved, which lays a foundation for the defect extraction and damage identification.