An Adaptive Image Denoising Algorithm Based on SVD and Energy Minimization
基于SVD和能量最小原则的图像自适应降噪算法

Based on SVD(Singular value decomposition) and the energy minimum principle,an adaptive image denoising algorithm is proposed in this paper.An energy model of image denoising in matrix norm is presented for restoring noisy image.By minimizing energe on the matrix norm,the singular number of the reconstructed image is adaptively determined.Comparing with the adaptive denoising algorithm based on compression ratio and SVD,it avoids calculating the function of image compression ratio and its knee point.Furthermore,it avoids solving nonlinear partial differential equation numerically comparing with the energy minimization denoising model of bounded variation,and so it never leads to local minimum in general.Our method can denoise rapidly and effectively.It can be also implemented easily in practice.Experimental results in the paper validate our proposed method.