Abstract—To sove the minimization problem of the energy method in image denoising, the diffusion process of the Euler-Lagrange could be decomposed to the diffusions along the tangential direction and normal direction. In order to denoising and retained the texture information of the real scene as possible, at the texture pixels, the diffusion along the tangential direction should be retained and the diffusion along the normal direction should be controlled. Therefore we propose a new denoising method with edge diffusion along the tangential direction directly. The performance of this method is tested by single images in this thesis.
Index Terms—Image denoising, energy method, edge diffusion, tangential direction, normal direction.
Shuang Liang, Shuli Wang, and Yu Wang are with School of Mathematical Sciences, Peking University, Beijing (e-mail: liangshuang12@pku.edu.cn, shuliw@pku.edu.cn,
wangyu_amo@pku.edu.cn).
[PDF]
Cite: Shuang Liang, Shuli Wang, and Yu Wang, "A New Method of Image Denoising Based on the Energy Method," International Journal of Information and Electronics Engineering vol. 7, no. 1, pp. 18-21, 2017.