【作者】 杨斯涵；

【导师】 王绪本；

【作者基本信息】 成都理工大学 ， 地球探测与信息技术， 2015， 博士

【摘要】 随着现代工业的不断发展,矿藏勘探程度的日益增加,目前我国浅地表已探明的矿产资源逐渐难以满足国民经济发展的需要,矿产资源勘查的关注对象已由以往的浅部矿转向了深部矿。重磁勘探对于实现地下不同地质体异常的有效分离,完成地质体分布、位置、产状及其地质含义的解释具有重要作用。同时,由于重磁勘探具有成本低、效率高、覆盖范围广、测量过程不受地域条件约束等特点,已经成为了地球物理勘探不可替代的重要方法。在重磁勘探中,由于实际测量得到的重磁数据是地下不同地质体引起的叠加异常,而这些异常是由地下不同埋深、不同规模地质体的综合反映,因此必须对这种叠加异常进行数据分析。目前,一方面可以通过对位场数据进行分离以获取不同埋深、不同规模地质体各自所产生的异常,进而分别予以研究;另一方面,可以对重磁异常数据进行边界增强,突出各地质体的边界,以实现快速的构造划分以及异常圈定,进而完成地质体异常范围的划分。论文以重磁位场数据处理技术研究为主线,在对传统异常分离、异常边界增强方法进行调研、试验、分析的基础上,讨论了传统方法的不足,并深入研究了以多尺度Curvelet分解理论为基础的粗、中、精不同尺度的异常分离方法,避免了传统“二分法”无法精细的在不同尺度上对地质体进行分离的缺陷。同时,深入研究并实现了基于偏微分方程的局部增强方法,提出了归一化索伯尔重磁异常边界增强方法以及形态学腐蚀与膨胀乘的边界增强方法,实现了在埋深大的弱异常边界增强的同时,达到不同深度异常边界均衡增强的目的。提出基于改进的Canny算子异常边界自动提取方法,使之更加适应实际数据中的连续边缘检测,为自动、快速、准确的异常边界定位提供了技术支持。具体而言,获得的研究成果主要包括以下:1、为了避免传统“二分法”难以精细的在多尺度上完成不同埋深与规模地质体叠加异常的有效分离问题,本文在深入研究Curvelet变换的多尺度分析性质及其去噪原理的基础上,得出了Curvelet多尺度分解后各层系数重构结果可以与地质体异常叠加分布的不同尺度分析相对应的结论。利用范围大、埋藏深的地质体通常表现为大尺度、平缓的异常分布,而范围小、埋藏浅的地质体通常具有小尺度、变化剧烈的异常分布的特点,通过模型试算,验证了curvelet变换的多尺度分析性质可以完成不同埋深与规模分布的叠加异常的多重分解目的。试验结果显示,该方法在对不同埋深与范围分布的重磁异常精细分离方面,较之切割法更具优势。进一步对实际重磁异常数据的处理结果表明,本方法能够有效的完成不同埋深与范围的叠加异常分离,为后续地质体物理含义的合理解释提供更加充分的信息。2、深入研究了重磁弱异常边界增强的方法,研究并实现了基于偏微分方程的局部增强方法,利用局部小区域的增强处理,缓解了采用全局操作所引起的增强后异常边缘发生形变的问题。为了同时完成深浅不同地质异常的边界识别,提出了归一化索伯尔的异常边界增强方法,采用nsob方法实现了强弱异常边界均衡增强的效果。提出了基于形态学腐蚀与膨胀乘的异常边界增强策略,采用形态学腐蚀与膨胀处理结果相乘的方法,并充分利用形态学膨胀方法对弱噪声的削弱功能以及形态学腐蚀方法对亮噪声的削弱功能,通过将两者相结合,达到了强弱异常边缘均衡增强的效果并降低了各类噪声的影响,使所获取的增强边缘更为连续。在对以上各类增强方法进行模型试算的基础上,对实际获取的重磁异常数据进行处理,获得了良好的异常边缘增强效果,为后续的异常边界可靠检测、异常边界定位、地质解释奠定基础。3、为了实现快速的构造划分以及异常圈定,需要实现异常的边界检测与定位。然而,由于重磁异常边缘在经过拉升或增强后,受采用的滤波器算子的大小限制所导致的地质体边界所产生的模糊或膨胀,难以准确表示各地质体真实边界。因此,提出了改进的canny算子异常边界自动提取方法,通过引入边界梯度的方向,利用边缘点所通常具有的统一、连续、缓变的梯度方向的特性以区别于噪声或增强所引起的边缘形变扩散,使其较之原canny方法更加适应实际数据中的连续边缘检测,为自动、快速、准确的异常边界定位提供了有力支持。本文在对前述各类方法进行模型试算的基础上,对实际采集的重磁异常数据进行了处理,并进一步结合处理结果对相关区域的地质构造特征及岩性分布特征进行了解释。重点针对云南永胜分水岭铜矿远景区开展的1:2万比例尺高精度磁法地面测量数据,采用本文所提出的Curvelet多层分解法,实现了该研究区域磁异常数据的多尺度位场分离。实际数据的处理结果显示,所提出的方法较之延拓法、切割法所得到的结果具有更为精细的多层次分析能力,证明了该方法在重磁位场分离与解释中具有良好应用前景。进一步,对该区域重磁数据进行了基于归一化索伯尔边界增强与形态学处理相结合的重磁异常边界增强处理,有效实现了强弱不同异常边界的均衡增强,并利用改进的Canny边缘检测方法,实现了地质体边界的精细提取。结合研究区的地质资料,完成了对研究区内地质构造特征以及岩性分布特征的解释。更多还原

【Abstract】 With the development of modern industry, the degree of mineral explorationis increasing and looking to the quality of shallow deposits has been difficult. It needs to provide strong technical support for deep mineral exploration. The gravity and magnetic method is an important branch of application of geological exploration and mineral for geophysical technology. It plays an important role for realizing the effective separation of underground geological bodiesanomaly and for completing geological body distribution, location, occurrenceand the geological interpretation. At the same time, because the gravity and magnetic method has the advantages of low cost, high efficiency, wide coverage, the measurement process without geographical constraints and other advantages, it has become an important and irreplaceable method for geophysical exploration.The actual gravity and magnetic data obtained is caused by the different underground geological factors of superimposed anomaly. It is the comprehensive reflection for different buried depth and scales of geological bodyunderground. For analyzing the superimposed anomaly, on the one hand, it can separate the potential field data and obtain the anomaly with different buried depth and scales for examing, and further research. On the other hand, it can enhance the edge of gravity and magnetic anomaly and realize the rapid tectonic division as well as anomaly, include completion of the geological anomaly range division.Based on the potential field data processing technology research, this thesis studies the traditional potential field separation and the anomaly edge enhancement technology. It studies the multi-scale curvelet decomposition theory deeply and realizes the separation of geological bodies with different scalesand depth. It avoided the shortcomings of the traditional "dichotomy" method based on the analysis on anomaly in different scales. For alleviating the obvious deformation problem of boundary shape after enhancement and achieving the effect of noise cancellation, the local enhancement method based on partial differential equations is promoted. To enhance the deep buried weak anomaly boundary and realize the enhancement equilibrium among gravity and magnetic anomaly with different depth, the normalized sobel edge enhancement method and the multiplication of morphology erosion and dilation methods are proposed to increase the weak and strong anomaly boundary at the same time, to achieve different depth anomaly boundary enhancement. In this thesis, the automatic extraction of anomaly boundary method based on canny operator is studied and an improved canny edge detection algorithm is proposed, which is more suitable to the actual data of continuous edge detection. To be specific, the research results mainly include the following:1、In order to avoid the question based on traditional "dichotomy" and to realize effective separation of different depth and scale of geological anomaly in multi scales, the multiscale transform analysis and the denoisingmethod based on Curvelet is studied. It can be concluded that the multi scale decomposition reconstruction results of each layer based on Curvelet transformation could corresponding to the different scales of geological anomalies. Because the geological body with large range and deeply buried usually has the large scale and gentle abnormal distribution, and the geological body with small and shallow buried usually has small and dramatic changes abnormal distribution. Through the model test, it can be concluded that the abnormal distribution with different depth and range could be decomposed based on the multi scale analysis ability of the Curvelet transformation. The results show that it is better than the result from the cutting method and can use to provide more adequate information for the rational interpretation of the physical meaning of the geological body.2、The weak gravity and magnetic anomalies enhancement method is studied through the partial differential equation based on local region processing. It ease the abnormal edge deformation problem after enhancement. In order to complete the identification of the anomaly boundary with different depth, the normalized sobel enhancement method is proposed. It can be enhancement balanced for the weak and strong boundary. For the same purpose, multiplication of morphology erosion and dilation methods are proposed to increase the weak and strong anomaly boundary at the same time. Because the morphology erosion could be used to remove the weak noise and the morphology dilation could be used to remove the strong noise. The influence of the noise is removed and the enhancement is balanced. The edge after enhancement is more continuous based on the method proposed in this thesis. The processing results based on the above enhancement method show that it can realize reliable detection of anomalous boundary.3、In order to realize the rapid tectonic division and the anomalydelineation, it need to achieve the boundary detection and location of abnormal. However, the edge of the geological boundary is usually fuzzy or expanded due to the influence of the size of filter after enhancement and it is difficult to accurately represent the real boundary around the Geological body. Therefore, a method of automatic extraction algorithm based on the canny operator is improved. By introducing the anomalous boundarygradient direction, and the feature that edge points usually has the characteristics of continuous and slow variation on gradient direction. The proposed method is more suitable to get the continuous edge detection result from the actual data and provides a powerful support for automatic, rapid, accuracy abnormal boundary localization.The gravity and magnetic anomaly data is processed based on the method proposed in this thesis, and it is explained further combined with the characteristics of geological structure and lithologic distribution characteristics of the area. Focusing on the scale of 1:2 million high-precision ground magnetic survey data in Yongsheng watershed in copper,Yunnan prospect. Using the method of curvelet decomposition of this paper proposed, the multi-scale field separation is realized. The results show that, the method proposed has better multi-level fine analysis ability than the continuation method and cutting method.It proves that the method has a good application prospect in the gravity and magnetic field separation and interpretation. Further, the regional gravity and magnetic data were processedwith normalized sobel edge enhancement and morphological processing method proposed in this thesis. It increase the weak and strong anomaly boundary at the same time.It realized the extraction of the geological body boundary based on the improved canny algorithm proposed in this thesis, and the characteristics of geological structure in the area and the lithologic distribution characteristics is interpretated.更多还原