基于偏微分方程和Gabor算法的图像细节特征保护方法

Methods of image detail characteristics protection based on partitial differential equation and Gabor algorithm

传统的数字图像恢复算法在图像恢复过程中会抹掉图像的局部特征如纹理或者细节特征,从而可能降低图像的可读性和可理解性。利用偏微分方程(PDE)和Gabor(PG算法)结合运用于图像恢复,不但能保护图像的边缘细节,而且还能有效地保护图像的局部特征,如纹理、有意义的小细节等反映图像重要信息的特征。将Gabor过滤器的图像特征识别特性融入PDE的图像恢复算法之中,对图像细节特征进行加权保护,再利用较简洁的基于局部能量残余的图像分区机制来控制图像光滑的强度,从而有效地解决传统PDE算法的"盲目性"。

Traditional digital image restoration algorithm in image restoration process would erase the local features of the image such as texture or detail characteristics, which may reduce the image”s readability and understandability. This article introduced the image restoration based on PDE and Gabor （ PG algorithm ） , which not only could protect the image edge details, but also could effectively protect the local characteristics of the image, and reflected the characteristics of important image information such as texture and meaningful small details. This article adopted image feature recognition characteristics of the Gabor filter into the PDE image restoration algorithm, protecting the characteristics of image detail weightedly, recycling more concise image partition mechanism based on local energy residual to control the intensity of the image smoothness, and solving the traditional ＂blindness＂ of the PDE algorithm effectively.