基于像素自适应邻域的纹理合成方法

针对传统像素点的纹理合成方法普遍存在对纹理样本适用范围比较窄的问题，提出了一种可以自动确定最佳匹配像素邻域大小的纹理合成方法.首先，利用方块编码原理提取样本纹理元，通过纹理元来确定最佳匹配像素邻域；然后，在像素匹配阶段，对像素匹配准则进行改进.通过比较样本像素和目标像素邻域的均值和方差来确定邻域的相似度，最终选择最优的合成像素.对比实验表明，该方法不仅有良好的合成效果，而且具有较广的适用范围，可以作为一种通用性的纹理合成方法.

Texture Synthesis Method Based on Pixel Adaptive Neighborhood

A novel method that can automatically confirm the best neighborhood size of pixel matching was proposed to improve the narrow application range of traditional pixel based texture synthesis. Firstly， texton was extracted using block truncation coding， and the best neighborhood size of pixel matching was confirmed by the texton. Then， the norm of pixel matching was improved in the matching stage. By comparing the mean values and variance of sample pixels and target pixels， neighborhood similarity and optimal pixel can be deternined. The results show that the proposed method can generate good synthesis and has a wide application range， and thus can be a universal texture synthesis technology.