地震资料解释系统是专门为解释地震资料的地质内容而设计的，是地质勘探领域不可缺少的解释工具。传统的地震解释都局限于二维的表达方式，对理解地质形态造成了困难。可视化技术的出现为其研究提供了新的途径，它能充分利用地震信息，以直观的三维图形对各种复杂数据场进行描述，并且解释速度快、精度高，成为研究油藏构造和隐蔽性油气藏储层的有效手段之一。 首先，对地震资料解释系统的现状、国内外发展方向以及课题的研究意义进行综述，分析了解释系统中的关键技术。 其次，从地震数据的研究入手，阐述了数据的采集过程、数据格式及其转换过程。通过读取地震数据，并借助OpenGL实现了地震资料的可视化，可以定制二维、三维立体显示等，使解释人员能直观地理解地质构造走向。在地层解释部分，研究地震层位解释的一般方法，并采用种子点引导的方法对三维层面进行自动追踪，结果表明该方法比传统方法解释速度快、精度高。在断层解释中，分析了相干技术及其三代相干算法，并应用改进的相干算法进行实验，效果图上断层清晰可见。对于断层两侧层位连续性的判定，采用了BP神经网络方法进行识别，训练结果与已知情况一致，可以使用该方法进行断层层位的推断。 最后，以VC++为开发工具，分别从地震数据管理、数据可视化、地震地层解释、断层解释四个方面对地震资料解释系统进行了设计与开发，实现了预期的功能，并对解释系统的进一步研究作以展望。

The system of seismic data interpretation is designed for explaining the geology details. it can quickly discover the geologic abnormal situation. The traditional two-dimensional method can not satisfy the interpretation need, it makes difficult to know the geological conditions. The visualization technique can fully use the 3D seismic information to describe each kind of complex data and relations by the 3D graphics which people understand easily, so the technique becomes one of effective methods to describe the structural problems of complex fault blocks and the storage of hiding oil field. Firstly, it is carried on the comprehensive summary to the present situation of this seismic data interpretation system and the domestic and foreign development direction. The meanings of this research are analyzed in details and several key technologies of system are researched. Secondly, the seismic data is analyzed, the data gathering process, the data format and the switching process has been introduced in this paper. Through read the seismic data and draws support from OpenGL to achieve the seismic data visualization, you can choose the method such as two dimensional and three dimensional displays and so on, so the personnel can understand the geologic structure trend intuitively. In the strata explanation part, the general method of the seismic explanation has been studied, and the automatic trace by the seed guidance has been carried out, the result demonstrated this method is better than the old one, because it has higher speed and more precision. In the fault interpretation, the coherence technique and its three generation algorithms has analyzed, and use the improvement algorithm carries on the experiment, the result of fault demonstration is obvious. Regarding the both sides position of the fault, used the BP neural network method to carry on the recognition, the training result is consistent with the known situation, so we can use this method to conclude the fault position. Finally, this interpretation system has been designed and developed from four modules by VC++ 6.0. They are seismic data management, data visualization, strata interpretation and fault interpretation. The anticipated function has been realized, and then the further research of interpretation system has been forecasted.