摘要
针对地质截面图内部的地层信息,提出了一种新颖的抗干扰性强的地层分割算法。该方法首先采用变差函数和置信度计算地质图中区域的方向角度,接着面向方向一致连续的地质图子区域设计基于方向流场引导的蛇模型分割方法;其次摒弃了基于梯度信息的图像方向结构求取方法,构造了融合Radon变换获取方向角度的Gabor滤波器,从而精确分割出不同地质层并修补了同一地质层内部的断裂处;最后通过在真实地质剖面图中进行的测试验证了该方法的有效性,其能获取更优的分割结果,从而对石油勘探和地质环境的三维重建起到推动作用。
For the stratum information about geological sectional view, a novel anti-interference stratum segmentation algorithm was proposed. Firstly the method used the variation functions and the confidence degree to measure the re- gional orientation angles, then based on the direction flow fields snake model, segmentation method was designed for the sub-regions with inconsistent directions. Secondly the method based on the gradient information to calculate image structure and orientation was abandoned, but image direction obtained by radon transform was integrated into the Gabor filter, thereby different geologic layers were accuratly extracted and internal gaps in the same geological layer were re- paired. Finally experimental results verify the dfectiveness of our proposed method which has a better texture segmen- tation results in the practical geological section map,and our method plays a role in promoting the oil exploration and geological environment three-dimensional reconstruction.
出处
《计算机科学》
CSCD
北大核心
2013年第1期277-281,297,共6页
Computer Science
基金
国家重点基础研究发展计划973项目(2011CB707904)
江苏省自然科学基金(BK2012418)
南京林业大学高学历人才基金(163070052)资助
关键词
纹理分割
方向流场
蛇模型
方向滤波器
Texture segmentation,Direction flow field, Snake model,Direction filter
