针对获取的红外热波图像存在加热不均、信噪比小、对比度低的特点，提出基于小波变换的独立分量分析法并将其应用于红外热波图像处理，该方法首先对红外图像进行小波分解，提取高频分量构建虚拟噪声通道，进而提高独立分量分析方法消噪性能。试验结果表明，该方法可以达到很好的消噪效果，使图像的缺陷显示更为清晰。

Aiming at the problems that the thermal wave image is subjected with the uneven heating, low signal-to-noise ratio and low defect contrast of the image, the paper proposes a kind of new ICA method based on wavelet transform and its application to the thermal wave image processing. The concept of virtual noise channel was introduced from high frequency component by wavelet transform and used to improve the performance in signal de-noising of ICA. The experiments indicate that good de-noised results can be obtained and the display of defects can be seen clearly with the method.