期刊文献+

基于地层切片的高精度边缘检测方法研究

Edge Detection Method Based on High-precision Strata Slicing
下载PDF
导出
摘要 为了克服常规边缘检测方法的精度低、计算效率低,且只能处理二维水平时间(或沿层)切片的缺陷,在常规的基于希尔伯特变换方法的基础上提出了一种新的基于地层切片的高精度边缘检测方法。该方法针对地层切片进行计算,扩展了沿层切片的应用范围,更符合地下实际的地质情况;并且采用时间域的希尔伯特变换进行计算,计算优于常规基于频率域的希尔伯特变换的方法。此外,在计算流程上针对地层切片数组的纵向和横向等2个方向同时进行处理:先采用二维保边去噪进行预处理,然后采用时间域希尔伯特变换进行边缘检测,不仅可以有效地压制噪声,而且可以进一步提高了计算精度。塔中某区的实际地震资料的应用效果表明,新方法的裂缝检测符合率高于常规的裂缝检测方法(如希尔伯特变换方法、相干方法、基于张量的方法等)。 In order to overcome there exists some defects of low precision,low computation,etc.in conventional edge detection method,and the methods can only process 2d horizontal time (or along the layer) slice.These defects,a new method for edge detection is put forward based on the formation of edge detection precision slicing on the basis of the conventional method based on Hilbert transform.The method calculates the slices of layer,which extends the range of applications of layer,and it is more realistic for geological conditions underground.And it uses time domain Hilbert transform,the calculate of which is superior to conventional Hilbert the transform based on the frequency domain.In addition,the calculation process of this method can process vertical and horizontal direction simultaneously for for the array of layer slice.First,the date using a two-dimensional edge-preserving denoising is pre-proceed,and then the time-domain Hilbert transform detecte edge,is used which not only can to effectively suppress noise,and can further improve the accuracy.The actual effect of the application of seismic data in the tazhong area show that a new method is put forward of crack detection coincidence rate is higher than the conventional method of crack detection (such as the Hilbert transform method,correlation method,the method based on tensor,etc).
出处 《科学技术与工程》 北大核心 2015年第1期168-173,共6页 Science Technology and Engineering
基金 国家自然科学基金项目(41274130)资助
关键词 希尔伯特变换 边缘检测 三维地震数据体 保边去噪 Hilbert transform edge detection 3D seismic data volume edge-preserving smoothing
  • 相关文献

参考文献19

二级参考文献78

共引文献246

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部