基于经验模式和全变分模型的图像去噪方法研究

文章利用经验模式分解算法的自适应性及高频噪声的强获取能力.对图像进行分解得到固有模式分量.再运用全变分模型对高频的固有模式分量进行去噪处理,处理后的分量与其他未进行处理的低频固有模式分量及剩余模式进行合成.采用峰值信噪比来比较处理后的加噪图像与处理前的加噪图像.结果表明该方法在提高峰值信噪比的同时,也使得图像的主观视觉效果更好.

Image Denoising Method Research Based on Empirical Mode Decomposition and Total Variation Mode

With the adaptability of empirical mode decomposition algorithm for and the strong access ability of the high frequency noise intrinsic mode function by decomposing the image.Then the total variation denoising model is adapted to deal with noise in high frequency intrinsic mode function,the component after processing and the low frequency mode component not handed and residual mode function compose a new image.Peak signal-to-noise ratio is adapted to compare the noise image before and after being disposed.The results indicate that the method not only increases peak signal-to-noise ratio,but also makes the image of subjective visual effect better.