基于盲解卷积的红外烟幕干扰图像去噪方法

Method for image restoration with iterative blind deconvolution under the infrared smoke interference

介绍了多频谱红外烟幕弹爆炸后,红外热像仪所采集图像产生噪声的原因及噪声模型;并根据烟幕干扰退化图像噪声的点扩展函数难以准确描述的特性,应用频域迭代盲解卷积算法进行红外图像去噪,根据算法原理并结合烟幕干扰图像的噪声特点,对算法进行了适当的改进。实验证明,该方法去噪效果明显,有效地克服了维纳滤波等传统图像去噪方法使图像变模糊的现象,并使烟幕干扰图像的有效区域得到增强。

This paper introduces the reasons of noise causing collection by electrooptical infrared imaging which inter- ferenced by muhispectral infrared bomb. It uses the iterative blind deconvolution in frequency domain to restore the image when the PSF is unknown. This paper properly improved the method of the image restoration based on the char- acteristics of infrared smoke interference images. The experiment proves that the methods of this paper have good effects of image restoring, it effectively avoided the blurring of images when wiping off the noises. Also it makes the regions of smoke screen for images to be enhanced.