首页 | 杂志简介 | 征稿简则 | 投稿指南 | 常见问题 | 名词解释 | 刊物订阅 | 联系我们 | English
无人机管线巡检中的测试桩识别
投稿时间:2019-09-02  修订日期:2019-09-09  点此下载全文
引用本文:胡进.无人机管线巡检中的测试桩识别[J].红外,2019,40(10):32~35
摘要点击次数: 138
全文下载次数: 65
作者单位E-mail
胡进 空军工程大学航空机务士官学校航空维修管理工程系 shiwen_79_1@163.com 
中文摘要:在无人机GPS信号丢失的情况下,测试桩的视觉辅助识别精度是影响油气管线自动巡检工作的关键因素。针对测试桩自动识别的精度问题,在分析测试桩及其周围地物背景目标特性的基础上,先用深度学习算法判断出测试桩被周围地物背景遮挡的情况。对于被遮挡的测试桩,采用不显著目标相对定位算法检测出测试桩的具体位置。最后通过现场采集的数据实验验证了文中算法的有效性。
中文关键词:管线巡检  测试桩  深度学习算法  不显著目标定位
 
Test-pile Detection in Pipeline Inspection by UAV
Abstract:In the case of GPS signal loss of the unmanned aerial vehicle (UAV), the auxiliary visual recognition accuracy of the test pile is a key factor affecting the automatic inspection of oil and gas pipelines. Aiming at the accuracy problem of automatic identification of test piles, based on the analysis of the background and target characteristics of the test piles and surrounding objects, a deep learning algorithm was used to determine whether the test piles were obscured by the surrounding objects. For obstructed test piles, the relative location algorithm of the insignificant target was used to detect the specific position of the test piles. Finally, the validity of the algorithm in this paper is verified by actual test scenarios experiments.
keywords:pipeline inspection  test-pile  deep learning algorithm  unsaliency target detectio
查看全文  HTML  查看/发表评论  下载PDF阅读器

版权所有:《红外》编辑部

北京勤云科技发展有限公司