基于图像识别技术的固井质量评价方法研究
DOI:
作者:
作者单位:

昆仑数智科技有限责任公司

作者简介:

通讯作者:

中图分类号:

基金项目:

中国石油天然气集团有限公司科研项目“钻完井及井下作业智能优化系统研发”(编号:2021DJ7401)。


Cementing quality evaluation method research by image recognition technology
Author:
Affiliation:

Kunlun Digital Technology Limited Liability Company

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    关于固井实施方法和固井质量工程探测方法研究颇多,而关于固井质量的图像解析却鲜有研究。本文就使用评价图进行固井质量评估问题,提出了用于聚焦图像解析范围的文本坐标位置参照算法,以及基于直方图相似度和梯度距离的图像解析算法。以图像解析算法和业务机理相结合的方式,从评价图中抽取出可评估固井质量的量化数据。所提出的图像解析算法能准确识别图片中的固井质量信息,两种算法的解析结果与实际结果误差率均在10%以内。本文所提出的方法可有效解决人工手动分析效率低且分析误差大的缺点,并对固井质量电子数据缺失仅留存固井质量评价图的情况提供了切实可行的解决方案,所提出的算法对于固井质量图片的深入挖掘具有较强的探索性意义。

    Abstract:

    There are many researches on cementing implementation methods and engineering detection methods of cementing quality, but the analysis of picture results on cementing quality is relatively rare. In this paper, a text coordinate position reference algorithm for focusing image analysis range and an image analysis algorithm based on histogram similarity and gradient distance are proposed to evaluate cement quality using evaluation image. A combination of image analysis algorithm and business mechanism is used to extract quantitative data from the evaluation image to evaluate cementing quality. Through comparison of model experiments, the proposed image analysis algorithm can accurately identify the cement quality information in the picture, and the error rate between the analytic results and the actual results is less than 10%. The method proposed in this paper can effectively solve the shortcomings of low efficiency and large error of manual analysis, and provides a practical solution to the situation that electronic data of cementing quality is missing and only the cementing quality evaluation image is kept. The proposed algorithm has strong exploratory significance for deep mining of cementing quality pictures.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2024-03-22
  • 最后修改日期:2024-06-06
  • 录用日期:2024-06-11
  • 在线发布日期:
  • 出版日期: