摘要
通过将边缘信息加入到传统TV模型中,改善扩散系数的收敛性,提出一种快速自动消除文字的改进TV模型。采用LoG算子和数学形态学对待处理图片中干扰文字区域进行检测与准确定位,将已确定的干扰文字区域作为改进TV模型的约束条件并进行求解即可有效消除文字的干扰。仿真结果表明,该方法对非纹理图片中干扰文字的检测和消除快速实用。
By adding edge information to the traditional TV model, the convergence of the diffusion coefficient is improved, and an improved TV model for quickly and automatically eliminating text interference is proposed. LoG operator and mathematical morphology are used to detect and accurately locate the interference text region in the image to be processed. By taking the determined interferential text region as the constraint condition of improving TV model and solving it, the interferential text can be effectively eliminated. The simulation results show that this method is fast and practical for detecting and eliminating interference texts in non-texture images.
引文
[1] 邵肖伟,刘政凯,宋璧.一种基于TV模型的自适应图像修复方法[J].电路与系统学报,2004,9(2):113-117.
[2] KROMMWEH J,MA J.Tetrolet shrinkage with anisotropic total variation minimization for image approximation[J].Signal Processing,2010,90(8):2529-2539.
[3] LARI H H,CHOLAML A.Curvelet-TV regularized Bregman iteration for seismic random noise attenuation [J].Journal of Applied Geophysics,2014,109:233-241.
[4] 王昊京,王建立,王鸣浩,等.采用双线性插值收缩的图像修复方法[J].光学精密工程,2010,18(5):1234-1241.
[5] 蔡占川,姚菲菲,唐泽圣.基于克里金插值法的图像修复[J].计算机辅助设计与图形学学报,2013,25(9):1281-1287.
[6] ALILOU V K,YAGHMAEE F.Non-texture image inpainting using histogram of oriented gradients [J].Journal of Visual Communication and Image Representation,2017(48):43-53.
[7] 许建楼,冯象初,郝岩.改进的TV-Stokes图像修复模型及其算法[J].电子与信息学报,2012,34(5):1142-1147.
[8] BARBU T.A novel variational framework for structural image completion[C].Joint International Conference Optim-Acemp,2017:815-820.
[9] 杜闪闪,韩超.基于TV模型的改进图像修复算法[J].激光与光电子学进展,2018,19(05):111-121.
[10] 王涛,王军锋,白青,等.改进的TV模型图像修复算法[J].计算机系统应用.2013,22(3):121-124+58.
[11] BOURDIN L,CRESSON J,GREFF I,et al.Variational integrator for fractional Euler–Lagrange equations[J].Applied Numerical Mathematics,2013(71):14-23.