自适应阈值的小波图像去噪

针对VisuShrink阈值和NormalShrink阈值的缺陷,提出了一种改进的自适应阈值图像去噪方法.根据不同的子带特性,定义了一个新的尺度参数方程,以确定适合各个尺度级的自适应最优阈值,并依此对图像进行去噪.实验结果表明,该方法可将每一尺度上的信号与噪声作最大分离,有效去除了白噪声,较好地保留了图像的细节信息,进一步提高了峰值信噪比,且没有增加时间复杂度,能用于实时处理.

Adaptive wavelet thresholding method for image denoising

An improved adaptive wavelet thresholding method for image denoising was proposed to overcome the limitation of Donoho's VisuShrink and Lakhwinder Kaur's NormalShrink. According to the different sub-band characteristics, a new scale parameter equation was defined based on Lakhwinder Kaur's NormalShrink threshold, which was employed to determine the optimal thresholds for each step scale. Experimental results on several testing images show that the proposed method separates signals from noise completely in each step scale and eliminates white Gaussian noise effectively. In addition, the method also preserves the detailed information of the original image well, obtain superior quality image and improves Peak Signal to Noise Ratio (PSNR). Furthermore, since this method can improve the efficiency of image denoising and doesn't increase time complexity, it could be applied in the real-time processing.