盲图像恢复中的二维盲均衡研究

Study of 2-Dimensional Blind Equalization in Blind Image Restoration

数字图像在传输过程中,由于受到二维信道的点扩展函数(PSF)的影响,会产生图像模糊问题。在对图像进行恢复时,需要消除PSF的影响,在许多实际应用中PSf的特性未知,此时图像的恢复就是盲图像恢复。该文提出了一种应用于盲图像恢复的二维盲均衡迭代算法,该算法是通信领域中应用的EVA盲均衡算法的一种二维扩展。仿真结果表明,该算法具有信噪比增益高、收敛速度快等特点,有较为广泛的应用前景。

During the transmission of digital image, the signal may be blurred by the Point Spread Function （PSF） of the 2-dimensional channel. On the restoration of the image, the influence of PSF should be eliminated without the knowledge of PSF in most real applications. It is called blind image restoration when the property of PSF is unknown. A new algorithm for blind image restoration via 2-dimensional blind equalization is proposed in this article. It is essentially a 2-dimensional expansion of EVA algorithm applied in the field of communication. Experimental results show that the algorithm works with rapid convergence speed and high improvement of SNR and has the prospect of wide applications.