根据小波变换和噪声信号地能量分布特性，提出了一种先用小波变换对含噪图像进行多尺度分解，求出各尺度小波变换高频系数地噪声方差和阈值，利用各尺度地阈值对高频系数进行处理，然后利用小波变换系数重构图像，实现图像降噪地方法。-According to wavelet transform and noise signal and energy distribution characteristics, a first wavelet transform of the noisy image multiscale decomposition, obtained the high-frequency-scale wavelet transform coefficients and noise variance and thresholds, using the scale to the threshold value of the deal with high-frequency coefficients, and then the use of wavelet transform coefficients reconstructed image, to realize image denoising methods.