基于软阈值的广义交叉验证准则(GCV)已运用于图像去噪中,但此方法对图像峰值信噪比(PSNR)的改善有限,不能有效保持细节.针对此问题,改进了广义交叉验证准则,且将平移不变应用于去噪.即先将原图像分别进行水平和垂直方向上的平移,然后将平移后的每幅图像变换到小波域,使用分块广义交叉验证准则对小波系数去噪后实行小波逆变换和反平移.最后将反平移后的各图像平均,即得恢复的图像.实验结果表明,该方法能有效恢复图像细节,图像的PSNR值和主观效果都有较大改善.

The denoising method based on GCV has been used in image denoising,but it is hard to improve PSNR greatly and can′t maintain the details of the image well.To solve the problem, an improved GCV combined with TI is put forward.Firstly shift the noisy image in horizontal and vertical directions respectively,and translate the shifted images into Wavelet-Domain and denoise with the method of GCV in blocks.Then translate the images by inverse discrete wavelet and shift them inversely.Finally the reconstructed image is obtained by average.Experiments show that this method reconstructs image details effectively and improves image quality both in PSNR and subjective sense.