沉积相分析对于岩性油气藏勘探与油藏研究十分重要。本文在地质研究的基础上总结出测井相模式,探索了用图像识别方法分析沉积相的可能性,使沉积相研究由取心资料这种常规的方法转变为利用测井曲线进行。识别算法包括训练和投票两个步骤,通过四种沉积相43个点的识别,结果是成功的。 Sedimentary facies analysis is very important for lithologic reservoir exploration and reservoir research. Based on the geological study, this paper summarizes the logging facies model and explores the possibility of using sedimentary facies to analyze sedimentary facies by using image recognition method. It makes the study of sedimentary facies transform from coring data to using logging data. The recognition algorithm consists of two steps, training and voting. Through the identification of 43 points of the four sedimentary facies, the result is successful.