基于改善逐行法恢复空间可变振动模糊图像

为了恢复振动造成的图像模糊,突出目标图像的特征。提出了一种考虑相邻行点扩散函数作用的逐行恢复算法,恢复空间可变振动模糊图像。首先分析了振动模糊图像频谱特征,即振动模糊图像频谱存在着对应零阶贝塞尔函数零点的暗纹。建立振动模糊图像退化模型,研究模糊图像每行受相邻行点扩散函数的影响。然后对一幅振动模糊图像进行二维傅里叶变换和Radon变换检测出其频谱上暗纹的方向,得到振动的方向θ。将模糊图像顺时针旋转θ至水平方向,再选取合适的窗口截取图像。最后采用考虑相邻行点扩散函数作用的逐行恢复算法恢复模糊图像。实验结果表明:该算法恢复图像效果优异,可获得细节清晰的图像。

Restore Space Variant Blurred Image of Vibration Based On Improved Iine-by-line Method

In order to restore blurred image caused by vibration, better highlight the characteristics of image. The improved Iine-by-line method considered effects of adjacent line point spread function is proposed. Resrore the space variant vibration blurred image. First, analyzed the charater of vibration blurred image in frequency domain, that is a set of negative ripples appears in specrums of blurred image corresponding to zero values of the zero order Bessel function. Established the vibration blur degraded model. Studied the each line of burre image affected by adjacent line point spread function. Then, some steps should be carried out to detect the derections of Bessel ripples in specrums of blurred image and to gain vibration derection angle ; they are 2D-Fourier transform, Radon transform and so on. Clockwise rotate the blurred image to horizontal derection by angle and then select the appropriate window to intercept the blurred image. Finally, use the improved Iine-by-line method considered effects of adjacent line point spread function to restore blurred image. The experimental results show that the algorithm has more better restore effect and gain much clearer details.