基于区域方差的纹理滤波算法

TEXTURE FILTERING ALGORITHM BASED ON REGION VARIANCE

纹理滤波器算法是图像处理领域重要的研究方向之一,具有广泛的研究意义和实际应用价值。一种类型纹理滤波器算法通过区分图像中包含的复杂纹理信息和显著结构信息来引导进行纹理滤波处理。如何区分图像中的纹理信息和结构信息成为了算法的关键。提出一种区分图像的复杂纹理信息和显著结构信息的方法。同时对于彩色的包含纹理信息的图像进行滤波处理时,引入了对比度保持的方法来生成灰度图像,用来生成指导纹理滤波的引导图。为了证明算法的有效性,给出了和一些算法进行比较的结果。另外,展示了算法在细节增强,JPEG压缩印迹去除和图像分割的应用效果。

Texture filter algorithm is one of the important research directions in image processing field,which has wide research significance and practical application value. One type of texture filter algorithm directs the texture filtering process by distinguishing between complex texture information and salient structure information contained in the image.How to distinguish the texture information and structure information in the image becomes the key of the algorithm.Aiming at this problem,a method to distinguish between complex texture information and salient structure information is proposed. Meanwhile,when the color image containing texture is filtered,the method of contrast preserving is introduced to generate the gray scale image,which is used to generate the guide map to guide the texture filtering. In order to prove the effectiveness of the algorithm,some algorithms are compared with the results. In addition,the application of the algorithm in detail enhancement,JPEG compression marking removal and image segmentation is showed.