This paper proposes a new wavelet domain denoising algorithm. In the results of conventional wavelet domain\r\ndenoising methods, ringing artifacts or wavelet-shaped noises are sometimes observed due to thresholding of small\r\nbut important coefficients or due to generation of large coefficients in flat areas. In this paper, nonlocal means filtering\r\nis applied to each subband of wavelet decomposition, which can keep small coefficients and does not generate\r\nunwanted large coefficients. Since the performance of nonlocal means filtering depends on the appropriate kernel\r\nbandwidth, we also propose a method to find global and local kernel bandwidth for each subband. In comparison\r\nwith conventional methods, the proposed method shows lower PSNR than BM3D when pseudo white Gaussian noise\r\nis added, but higher PSNR than the spatial nonlocal means filtering and wavelet thresholding methods. For the\r\nmixture noise or Poisson noise, which may better explain the real noise from camera sensors, the proposed method\r\nshows better or comparable results than the state-of-the-art methods. Also, it is believed that the proposed method\r\nshows better subjective quality for the noisy images captured in the low-illumination conditions.
Loading....