图像的纹理标准性度量及其应用

通过分析纹理合成中子块参数对合成速度及质量的影响，发现对于一类纹理图像，不依赖于参数的选择即可快速高效地进行纹理合成。选择子块灰度平均值作为度量指标，根据纹理标准性强的图像应具有的特征，提出了一种新的计算图像纹理标准性系数的算法。结合大量计算结果，界定了一般图像、强标准性纹理以及弱标准性纹理的分类标准。并将其应用到纹理合成当中，对强标准性纹理图像的合成采用大尺度子块及零搜索的合成方法，提高了合成速度的同时保持合成质量不变。

Measurement and application of image's texture standard property

By analyzing the effect of sub-block parameters on speed and quality of texture synthesis,it has been found that a class of texture images which don＇t depend on the parameters can obtain rapid and effective synthesis.This paper chooses mean gray value of sub-block as a measure,proposes a new algorithm to calculate the coefficient of texture standard property.According to a large number of statistical results, a criterion of distinction of strong standard textures and weak standard textures and general images has been given, and it is applied to texture synthesis.Large scale sub-block and zero search is used to synthesize strong standard textures.The results show that even compared to the global search methods, it can obtain rapid and effective results.