基于领域自定义语言的探地数据可视化研究

Visualization has strong advantages and good potentials in complex data analytics, such as seismic data, ground penetrating radar data, oil reservoir data, etc. They provide a visual encoding to get insight into revealing attribute information presented in the data. It assists users perform effective analytics by means of various human-computer interactions. The existing illustrative visualization methods for seismic data still have some weakness in the interactive exploration, for example, the illustrative process is tedious, the control accuracy is not fine, the whole process lacks of customizability, etc. In this proposal, we present a Domain Specific Language (DSL)-based method to solve these problems in seismic data visualization, which helps to enhance the customizability in interaction and allows re-editing and re-using the illustrative animation. Specifically, DSL is utilized to strengthen the customizability on algorithm selection and visual style formatting. A visual style brush will be carefully designed to support users to customize various visualization techniques and explanatory style. It builds a framework to support the parameterization representation and formal language definition for spatial domain interactions, which provides various display schemes and further focus + context exploration in local domain. Furthermore, it enables to build illustrative animation for seismic volume data, which makes the generation of interpretation animation more flexible, editable, and reusable. Finally, the evaluation will be conducted together with the domain experts who engaging in oil/gas searching many years.

可视化在石油勘探领域的探地数据、油藏数据等复杂数据分析和表现方面具有独特优势，可以将数据中内蕴的属性信息进行视觉呈现，借助丰富的人机交互手段，辅助用户进行有效的分析。基于现有的探地数据可视化方法在交互方面效率不高，控制精度不细，无法灵活定制等缺点，本研究拟采用领域自定义语言对探地数据解释可视化进行交互增强，以增强交互的可定制性与解释动画的可编辑性和可复用性。具体而言，拟使用领域自定义语言增强二维沉积相解释可视化的算法与风格定制能力，使用风格格式刷以支持用户定制不同的可视化方法及解释风格；提供三维地震体空间域交互的参数化表示及形式化语言定义，以支持多种复杂的显示方式以及针对现有当前结果进行进一步的焦点上下文式的局部探索；构建三维地震体的解释动画，以提供更加灵活的、可编辑的、可重用的解释可视化动画。针对不同的交互增强，分别为石油勘探领域的领域专家或技术人员设计不同的测试用例进行综合比较与评价。

数据可视化与可视分析在石油天然气勘探领域的探地数据、油藏数据等复杂数据分析和表现方面具有独特优势，可以将数据中内蕴的属性信息进行视觉呈现，借助丰富的人机交互手段，辅助用户进行有效的分析。基于现有的探地数据可视化方法在交互方面效率不高，控制精度不细，无法灵活定制等缺点，本研究拟采用XML等领域专家学习曲线低的定制语言对探地数据解释可视化进行交互增强，以增强交互的可定制性。主要是增强二维沉积相解释可视化的算法的定制能力，提供三维地震体空间域交互的参数化表示及形式化语言定义。借助该项目的资助，我们的成果在探地数据可视化研究领域起到了一定的推动作用。