Pattern recognition in electrical images - digital image analysis with special focus on Gas Hydrate Leg 204

电图像中的模式识别 - 数字图像分析，特别关注天然气水合物 Leg 204

• 批准号: 5434755
• 负责人: Professor Dr. Christoph Clauser
• 金额: --
• 依托单位: Informatik 5 - Lehr- und Forschungsgebiet Life Science Informatik
• 依托单位国家: 德国
• 项目类别: Infrastructure Priority Programmes
• 财政年份: 2004
• 资助国家: 德国
• 起止时间: 2003-12-31 至 2005-12-31
• 项目状态: 已结题
• 来源: https://gepris.dfg.de/gepris/projekt/5434755?language=en
• 关键词: Pattern; recognition; electrical; images; digital

Electrical images are widely used in oil industry for borehole inspection and reservoir characterisation. They consist of artifically colour-coded micro-resistivity measurements on the borehole wall. In ODP these are commonly used to study the internal structure of the oceanic crust. At present, the reconstruction of lithology from electrical images is purely based on visual inspection and subjective interpretation. Our objectives are to (1) develop methods for an object- and texture-based pattern recognition of electrical images; (2) develop methods for an automatic classification of rocks; (3) apply this methodology to study the occurrence and characteristics of gas hydrates at ODP Leg 204. Avaible methods from seismic data interpretation and biomedical imaging will be adapted for use with resistive images. In particular, we will implement an algorithm for automatic object identification for a study of the internal rock morphology with respect to sediments drilled at ODP Leg 204. This way, resistivity anomalies caused by gas hydrates will localised them in electrical images. Their proportion, size, and shape will be quantified in order to analyse anisotropy and spatial heterogeneity of identified gas hydrates. Finally, we will correlate the 1D information from all available boreholes and interpolate them into a 2D map of the gas hydrates distribution at Leg 204.

电图像在石油行业中广泛用于钻孔检查和储层表征。它们由钻孔壁上的人工颜色编码的微克抗性测量。在ODP中，这些通常用于研究海洋外壳的内部结构。目前，从电图像重建岩性的纯粹是基于视觉检查和主观解释。我们的目标是（1）开发用于基于对象和纹理的电气图像的模式识别的方法； （2）开发自动分类岩石的方法； （3）将此方法应用于ODP腿204的气体水合物的发生和特征。可抗地震数据解释的可行方法和生物医学成像将适用于​​电阻图像。特别是，我们将实施一种用于自动对象识别的算法，以研究内部岩石形态在ODP腿204处钻孔的沉积物。这样，通过气体水合物引起的电阻率异常将将它们定位在电图像中。它们的比例，大小和形状将被量化，以分析已鉴定的气体水合物的各向异性和空间异质性。最后，我们将从所有可用的钻孔中关联一维信息，并将它们插入腿部204处气体水合分布的2D图中。