中文摘要：

微电阻率扫描成像测井（FMI）可提供井筒周围清晰、直观的岩层图像。在沉积微相分析中,用岩心资料和常规测井资料对FMI测井图像标定后进行沉积模式识别,能增加分析结果的可靠性。以往用FMI测井图像沉积微相分析时,需要大量人工参与,解释效率低。基于计算机图形学算法与神经网络分类器原理,提出了一种直接由FMI图像自动识别沉积微相的方法,与传统沉积微相识别方法相比提高了解释效率并减少了人为因素的干预。实践表明用该方法识别沉积微相,可靠性和识别效率得到很大提高。

英文摘要：

FMI（formation micro-resistivity imager） log technology provides a clear and intuitionistic image of stratum around borehole wall.The core data and conventional logging data are usually used to calibrate the FMI image for sedimentary facies identification which can improve the reliability of the analyzed result.The traditional sedimentary facies identification method based on imaging logging requires lots of manual calculations,so it has a low interpretation efficiency.Based on computer graphics algorithm and neural networks classifier,proposed is a method for sedimentary facies auto identification from FMI images.Compared with traditional sedimentary facies identification method,this method can improve interpretation efficiency and reduce artificial influence.The log practice shows that the method for sedimentary facies identification can highly improve interpretation reliability and efficiency.