遥感模糊降质图像的扩散恢复算法研究及实现

图像恢复是图像处理中非常重要的一个方面,其方法是通过相应的退化模型和算法,对退化降质或受到噪声污染的图像进行重建,恢复原来的图像。作者介绍了利用传导扩散及反扩散模型,对因物体运动、大气湍流的扰动,拍摄时的散焦等引起的模糊降质图像的恢复算法,并给出了Laplace算子离散方法以及实现程序,提出了图像恢复的递推、迭代模型,基本解决了图像“盲恢复处理”的问题。对所述算法进行了全面的图像恢复测试、比较,证明算法具有较大的适应范围,可应用于勘探、遥感、司法刑侦、无损探测等领域的图像恢复处理。

RESEARCH AND IMPLEMENTATION OF DIFFUSION RESTORATION ALGORITHM ABOUT BLURRED DROP-QUALITY IMAGE OF REMOTE SENSING

It is very important respect that image restoration in picture processing. The method is to (recreate) and restore the degenerated and blurred drop-quality image to the original image with suitable (degeneration) model and algorithm. This paper shows the method how to use conduction model of diffusion and inverse diffusion to restore the blurred image and gives the method of Laplace algorithm, restoration examples and implementation (program). The paper introduces the model of resuming image gradually and solves the problem of "blind restoration process" in image processing basically. The paper tests an overall (restoration) situation of blurred drop-quality image and proves the algorithm very good. This algorithm can be used in exploration, remote sensing, administration of justice, survey of not dampened, and so on.