综合热扩散和全变分模型的自适应图像去噪模型

摘要

讨论了各向同性热扩散(TD)方程模型和各向异性全变分扩散(TV)方程模型对图像去噪的影响。在此基础上提出了一个改进的自适应混合模型，所提出的模型根据图像的信息能够自适应每个区域。该模型进行扩散图像的更加平坦区域，而较少的扩散在图像的边缘，得到了较好的去噪效果，同时保持边缘。比较了这三个模型对图像去噪的效果。数值实验结果表明，改进的自适应混合模型具有更好的去噪效果。%The paper mainly discussed the influence of the TD (isotropic) and TV (anisotropic) equation models on image denoising. And on the basis of this research ,the paper proposes an improved model, named adaptive hybrid model .The newly proposed model can be adapt to each area according to the information of image. The model diffusesthe smooth areas of the image, but diffuses less on the edge of the image, for which it reaches to better effect of denoising. And maintain the edge in the meantime. After comparing the three kinds of with each other, the numerical results of the experiment indicates that, the improved adaptive hybrid model has better effect of denoising.