非局部平均是当前一种新兴而有效的图像去噪方法。为了能充分利用数字图像局部几何结构的自相似性,同时由于结构张量可有效刻画数字图像的局部几何结构特征,进而提出了基于结构张量相似性度量的非局部平均去噪算法。实验结果验证了该算法抑制噪声的有效性,同时能很好地保持边缘等细节特征,峰值信噪比得到有效提高。
Non-Local Means(NL-Means) algorithm is an emerging and effective image denoising method.In order to sufficiently use the self-similarity of the image local geometric structures,this paper proposes a NL-Means image denoising algorithm based on structure tensor for similarity measure,which is capable of describing the self-similarity of local geometric structures in the image.The experimental results demonstrate the effectiveness of the algorithm,also show that the method has advantages for both edge preserving and PSNR value.