文摘
|
针对无人机巡检中输电走廊施工车辆入侵这一潜在危险事件,提出一种基于图像处理和机器学习的施工车辆自动检测算法。首先,对于无人机收集到的图像进行相应的预处理;其次,根据施工车辆的颜色、直线结构等特征,给出2种施工车辆目标区域提取方法,有效缩小了识别范围;最后,选择HOG特征与支持向量机(SVM)结合的方式,给出一种施工车辆识别方法。实验表明,算法能够在输电走廊下复杂场景中检测到施工车辆的存在,有较好的准确率。 |
其他语种文摘
|
In order to solve the problem of construction vehicles under the transmission corridor in UAV inspection, an automatic detection algorithm for construction vehicles based on image processing and machine learning is proposed. First of all, the image collected by the UAV is preprocessed; secondly, according to the color and linear structure features of construction vehicles, two extraction methods of construction vehicles are given, effectively narrowing the range of recognition; finally, this paper chooses the method based on HOG feature and support vector machine (SVM), and gives a construction vehicle identification method. The experimental results show that the proposed algorithm can detect the presence of construction vehicles in complex scenes under the transmission corridor, and has good accuracy. |
来源
|
控制工程
,2019,26(2):246-250 【扩展库】
|
DOI
|
10.14107/j.cnki.kzgc.161172
|
关键词
|
施工车辆检测
;
无人机巡检
;
Hough变换
;
HOG特征
;
支持向量机
|
地址
|
1.
中国科学院自动化研究所, 复杂系统管理与控制国家重点实验室, 北京, 100080
2.
国家电网吕梁供电公司, 山西, 吕梁, 033000
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
1671-7848 |
学科
|
自动化技术、计算机技术 |
基金
|
国家自然科学基金项目
|
文献收藏号
|
CSCD:6428595
|