自适应聚类Hough变换及地震断层检测

发布时间：2018-03-07 02:29

本文选题：Hough变换　切入点：地震图像　出处：《高技术通讯》2017年03期 　论文类型：期刊论文

【摘要】：针对现有基于Hough变换的地震断层检测方法只能检测单个断层,不能准确检测多个断层的不足,提出了一种基于自适应聚类Hough变换的地震断层检测方法。该方法首先对预处理后的地震相干图像进行边缘检测并对边缘图像进行Hough变换以检测出边缘图像中的线段,然后根据倾斜角和位置信息对线段进行自适应聚类以获得更完整的线段,最后根据初始地震图像对完整线段中的各点进行调整以获得准确、平滑的断层。为验证该方法的有效性,在实际地震图像上进行了对比实验。实验结果表明,该方法可正确检测地震图像中的多个断层,正确率达到90%以上,与现有方法相比,峰值信噪比提高了约10%。
[Abstract]:The existing seismic fault detection methods based on Hough transform can only detect a single fault, but can not accurately detect multiple faults. An adaptive clustering Hough transform based seismic fault detection method is proposed, which firstly detects the edge of the pre-processed seismic coherent image and the edge image by Hough transform to detect the line segments in the edge image. Then according to the slope angle and position information, the line segment is clustered adaptively to obtain a more complete line segment. Finally, the points in the complete line segment are adjusted according to the initial seismic image to obtain the accuracy. In order to verify the effectiveness of this method, a comparative experiment is carried out on the actual seismic images. The experimental results show that the method can correctly detect multiple faults in seismic images, and the accuracy rate is over 90%, compared with the existing methods, Peak signal-to-noise ratio (PSNR) is increased by about 10%.
【作者单位】：北京工业大学计算机学院;中国科学院地质与地球物理研究所;
【基金】：国家自然科学基金(61501008) 北京市自然科学基金(4162007)资助项目
【分类号】：P631.4;TP391.41

【相似文献】