视觉结构相似度地震图像质量评价模型研究
发布时间:2019-01-05 06:31
【摘要】:地震图像质量的评价通常采用定性与定量评价标准来衡量。文中提出了一种新的地震数据处理量化评价模型——基于全参考型视觉结构相似度(Seismic Data Structural Similarity,SDSS)的地震图像质量模型,通过计算资料处理前、后图像之间的能量强度测度、对比度测度与反射结构相似度测度,对图像的质量进行综合评价,用量化指标体现处理前、后图像的变化趋势。算法的数值模拟和实际资料结果表明,该方法与传统的评价方法相比,易于理解且计算简单,能凸显地震资料变化的差异,极大地提高了地震图像客观评价结果与人为主观感知的一致性。
[Abstract]:The evaluation of seismic image quality is usually measured by qualitative and quantitative evaluation criteria. In this paper, a new quantitative evaluation model for seismic data processing is proposed, which is based on the full reference visual structure similarity degree (Seismic Data Structural Similarity,SDSS). The energy intensity measurement between the images before and after the data processing is calculated. The contrast measure and the similarity measure of reflection structure are used to evaluate the image quality synthetically. The change trend of the image before and after processing is reflected by the quantization index. The results of numerical simulation and practical data show that the method is easy to understand and easy to calculate compared with the traditional evaluation method, and it can highlight the difference of seismic data change. It greatly improves the consistency between objective evaluation results of seismic images and human subjective perception.
【作者单位】: 西南科技大学环境与资源学院;中国石化西北油田分公司勘探开发研究院;
【基金】:国家自然科学基金项目(41204068)资助
【分类号】:P631.4
本文编号:2401411
[Abstract]:The evaluation of seismic image quality is usually measured by qualitative and quantitative evaluation criteria. In this paper, a new quantitative evaluation model for seismic data processing is proposed, which is based on the full reference visual structure similarity degree (Seismic Data Structural Similarity,SDSS). The energy intensity measurement between the images before and after the data processing is calculated. The contrast measure and the similarity measure of reflection structure are used to evaluate the image quality synthetically. The change trend of the image before and after processing is reflected by the quantization index. The results of numerical simulation and practical data show that the method is easy to understand and easy to calculate compared with the traditional evaluation method, and it can highlight the difference of seismic data change. It greatly improves the consistency between objective evaluation results of seismic images and human subjective perception.
【作者单位】: 西南科技大学环境与资源学院;中国石化西北油田分公司勘探开发研究院;
【基金】:国家自然科学基金项目(41204068)资助
【分类号】:P631.4
【相似文献】
相关期刊论文 前2条
1 谢江涛;雷英成;杜浩坤;;纹理模型回归分析在地震图像增强中的应用[J];长江大学学报(自然科学版);2011年05期
2 ;[J];;年期
相关会议论文 前1条
1 郭宇宏;王伟;高静怀;;基于线性反演方法的地震图像拉平方法[A];中国地球物理学会第二十七届年会论文集[C];2011年
相关硕士学位论文 前7条
1 谭小勰;三维图像分类方法及应用研究[D];电子科技大学;2016年
2 聂明明;基于模糊集的地震图像增强方法研究与实现[D];大庆石油学院;2010年
3 连涛;二维/三维地震图像的融合显示技术研究[D];电子科技大学;2013年
4 王春艳;地震图像序列非线性相干增强扩散滤波技术研究[D];大庆石油学院;2008年
5 祖国峰;基于小波变换的地震图像增强方法研究与应用[D];东北石油大学;2011年
6 陆华;多结构自适应形态滤波及其在地震图像处理中的应用[D];成都理工大学;2012年
7 彭军军;地震图像的纹理特征提取及分类[D];成都理工大学;2011年
,本文编号:2401411
本文链接:https://www.wllwen.com/kejilunwen/diqiudizhi/2401411.html