基于GPU的多属性数据快速融合渲染研究

摘要

由于传统的渲染技术是使用CPU进行数据体颜色计算或融合处理的，这种技术对大规模数据体进行渲染时效率低、时间长，针对这种情况提出一种采用GPU进行数据体颜色计算和融合处理的方法。该方法充分利用GPU强大的并行处理能力，将待渲染的数据以纹理形式提交给GPU，由GPU进行必要的颜色插值和融合处理后直接渲染。实验结果表明，该方法能够将多种属性融为一体，有机地结合了各属性的优点，能对油气储层进行综合评价，提高储层分析和解释的准确度，并且使用了硬件加速功能，渲染速度快。%In the traditional rendering technology, the calculation of data volume or the fusion process is realized with CPU, which is inefficiency and time-consuming to render massive data volume. To solve this problem, we propose an approach which us-es GPU to deal with color calculation and data fusion. This method takes advantage of the parallel processing capabilities of GPU. Data will be submitted to GPU in the form of textures, and then GPU deals with necessary color calculation and data fusion and fi-nally renders them. Experimental results show that the method can make integration of multi-attribute data, organically combines the advantages of each attribute, carries out a comprehensive evaluation of oil and gas reservoirs, and improves the accuracy of reservoir analysis and interpretation. The rendering is accelerated with GPU.