一种空间变化PSF图像分块复原的拼接方法

A Splicing Method of Sectioned Restoration Algorithm for Images with Space-Variant Point Spread Function

介绍了一种退化函数随空间变化图像的分块复原和拼接方法,首先以等晕区为依据对退化函数随空间变化的图像进行分块并向外延伸一定区域,然后使用Lucy-Richard算法对每一个图像块进行图像复原,在去除各复原图像块包含明显振铃波纹的部分边缘区域之后,将相邻图像块剩余的重叠区像素以其到等晕区边界的距离构成的加权系数进行叠加,拼接得到完整的图像。该方法能够消除退化函数随空间变化的图像分块复原处理时产生的振铃波纹,在诸如卫星遥感、飞机航拍图像复原,医学图像处理等众多领域均可广泛应用。

A sectioned restoration algorithm for images with space-variant point spread function and the according splicing method are introduced. Firstly, the blurred image is divided to sub-frame sections based on isoplanatic region and extended outward region. Then, the sub-sections are restored by Lucy-Richard algorithm. After removing the boundary regions which contain obvious ringing artifacts, the pixels in the rest of overlapping region between adjacent sections are weight stacked. The weight coefficients are related to the distances between the pixels and the boundary. At last, the sub-frames are spliced together to construct the composite full-image. The method is effective for eliminating the ringing artifacts coupled with sectioned image restoration with space-variant degraded function. It can be used in image restoration of remote-sensing with satellite and aerocraft, and also medical image processing.