摘要
随着PDC钻头的推广使用,传统的岩屑录井方法已难以分析如此细小的岩屑。基于数字图像处理技术对岩屑分类识别技术进行研究,首先利用和差直方图统计方法对岩屑进行纹理特征分析和提取,接着运用贝叶斯分类器进行分类识别,实验结果表明,此统计特征提取方法对于大部分岩屑可以很好地获取其主要特征,并最终取得理想的识别结果。此技术的发展将提高现场录井人员的工作效率和识别准确率。
Because of the application of Polycrystalline Diamond Compact Bit,the traditional cutting logging methods are facing many difficulties,and they are no longer applicable to describe so small cuttings.So researches are based on the numerical image processing technology.First,analyses and extracts the texture features of cuttings with sum and difference histograms.Then,recognizes it by Bayes classifier.The result suggests it is useful to acquire the main features for most cuttings and has a good result of recognition.The improvement of this technology will raise rapidly the work efficiency and the accuracy of recognition for site geologists.
出处
《中国海洋大学学报（自然科学版）》
CAS
CSCD
北大核心
2011年第3期99-104,共6页
Periodical of Ocean University of China
基金
国家高技术研究发展计划项目(2002AA615170)
中国石化胜利油田有限公司技术开发项目资助
关键词
岩屑
岩性
和差直方图
识别
cuttings
lithology
sum and difference histograms
recognition
