中国安全科学学报 ›› 2024, Vol. 34 ›› Issue (S1): 165-171.doi: 10.16265/j.cnki.issn1003-3033.2024.S1.0028

• 安全工程技术 • 上一篇    下一篇

改进型Harris角点算法的边坡滑动智能监控

陆鹏1(), 王占宝2, 余枚姣3, 张仁永3, 吴燕清1   

  1. 1 重庆大学 资源与安全学院, 重庆 400044
    2 国家能源投资集团有限公司 安全环保监察部, 北京 100011
    3 重庆工程学院 电子信息学院, 重庆 400056
  • 收稿日期:2024-03-12 修回日期:2024-05-15 出版日期:2024-06-30
  • 作者简介:

    陆 鹏 (1987—),男,四川武胜人,博士研究生,工程师,主要研究方向为物探相关数据采集、虚拟仪器技术、图像处理等。E-mail:

    王占宝, 工程师;

    余枚姣, 讲师;

    张仁永, 副教授;

    吴燕清,教授。

  • 基金资助:
    重庆市教委科学技术研究计划青年项目(KJQN202101910)

Intelligent monitoring of slope sliding with improved Harris corner point algorithm

LU Peng1(), WANG Zhanbao2, YU Meijiao3, ZHANG Renyong3, WU Yanqing1   

  1. 1 School of Resources and Safety Engineering, Chongqing University, Chongqing 400044, China
    2 Safety and Environmental Protection Supervision Department, CHN Energy Investment Group, Beijing 100011, China
    3 School of Electronic Information, Chongqing Institute of Engineering, Chongqing 400056, China
  • Received:2024-03-12 Revised:2024-05-15 Published:2024-06-30

摘要:

为减少边坡滑坡及其引发的地质灾害对财产安全的严重威胁,基于改进的Harris角点检测技术的匹配算法,提出一种新型边坡滑动监测方法。首先,通过优化Harris角点检测构建预测边坡滑动位移的图像处理模型;然后,基于校准的模板,监测角点位置的变化,从而精确估计滑动位移并追踪其轨迹,进而有效监控边坡的移动情况;最后,搭建快速的模块化试验环境,开展边坡滑动情况监测。结果表明:该方法降低了边坡监测识别中的伪角点生成,提高了角点检测的准确性和效率,能有效记录目标物体在水平、垂直和倾斜方向上的微小滑动,形成滑动轨迹追踪结果。

关键词: 角点检测, 边坡滑动监测, 特征点, 图像处理, 滑动轨迹

Abstract:

In order to reduce the serious threat of slope sliding and the geologic disasters they cause to property safety, a novel slope sliding monitoring method was proposed by using a matching algorithm based on the improved Harris corner point detection technique. First, an image processing model for predicting slope sliding displacement was constructed by optimizing the Harris corner point detection; then, based on the calibrated template, the changes in the location of the corner points were monitored to accurately estimate the sliding displacement and track its trajectory, so as to effectively monitor the movement of the slope; finally, a modularized test environment was constructed to monitor slope sliding conditions. The results show that the method reduces the generation of pseudo-corner points in slope monitoring and identification, improves the accuracy and efficiency of corner point detection, and effectively records the tiny sliding of the target in the horizontal, vertical, and inclined directions to form trajectory tracking results.

Key words: corner detection, slope sliding monitoring, feature points, image processing, sliding trajectory

中图分类号: