基于图像纹理合成的特征可调控的数字岩心重构方法

Feature-adjustable digital core reconstruction method based on image texture synthesis

为弥补现有的数字岩心重构方法难以进行特征调控的不足,文章提出一种新的基于特征子图且能够自由调控特征比例的数字岩心重构方法。该方法从训练图像中挑选出有代表性的特征子图,根据事先设置的比例进行匹配块挑选,再利用Graph Cut算法合成重叠区域,一直到全图完成。结果表明该方法能够提高合成速度,控制重构岩心图像中矿石、有机质等组分比例,并且能够生成新的模式,具有良好的应用前景。

The existing digital core reconstruction method is difficult to control the feature.To compensate for the shortcomings of the existing methods,this paper proposes a new digital core reconstruction method based on feature subgraph and capable of freely regulating the feature proportion.The method selects a representative feature subgraph from the training image,selects the matching block according to the preset proportion,and then uses the Graph Cut algorithm to synthesize the overlapping region until the whole image is completed.The results show that the method can improve the synthesis speed,control the proportion of ore,organic matter and other components in the reconstructed core image,and can generate new patterns,which has a good application prospect.