基于无人机视觉的道路违法搭建检测  

Illegal Construction Detection of Road Based on Unmanned Aerial Vehicle Vision

在线阅读下载全文

作  者:张晨 刘宁钟[1] ZHANG Chen;LIU Ning-zhong(School of Computer Science and Technology,Nanjing University of Aeronautics and Astronautics,Nanjing 211106,China)

机构地区:[1]南京航空航天大学计算机科学与技术学院,江苏南京211106

出  处:《计算机技术与发展》2018年第7期140-143,154,共5页Computer Technology and Development

基  金:国家自然科学基金(61375021);江苏省2016年交通科技与成果转化项目

摘  要:为保证良好的路域环境,保障道路安全畅通,对道路周围违法违规堆积物或搭建物的检测工作就显得尤为重要。传统的人工检测方法不仅工作量大、投入成本高,而且工作效率低、检测准确度不高,因此对道路违法搭建物的自动化智能检测识别方式迫在眉睫。提出了无人机飞过相同路段两次进行拍摄,通过数字图像处理和机器视觉技术对路段中发生明显变化的区域进行自动识别,以检测疑似违法搭建物的方法。对于两次拍摄的时间和空间的不匹配性,采用了直方图规定化和SURF特征点匹配变换的方法以使得两次拍摄图像具有相同模式。通过边缘检测和形态学滤波等方法对道路两侧监控区域实现自动识别。对于客观拍摄条件的不稳定性和景物的时延变化带来的干扰具有很好的鲁棒性,同时对于明显变化的图像景物区域的检测准确率高。实现了对道路违法搭建物的稳定、高效、准确的自动检测。To ensure a good road environment and the safety of roads,it is particularly important to detect the illegal deposits or structures around the road.The traditional manual detection method not only requires a lot of work and cost,but also results in lowwork efficiency and detection accuracy.So the intelligent detection and identification of the road illegal structures is imminent.To detect suspected illegal structures,we propose a method that uses UAV flying over the same road twice to shoot and then automatically identifies areas of significant changes in the road by digital image processing and machine vision technology.There is a mismatch between time and space for two shots,and the method of histogram normalization and SURF feature point matching transformation is adopted so that the two captured images have the same pattern.The edge detection and morphological filtering and other methods on both sides of the road are used to achieve automatic identification area.This method not only achieves a great robustness for the instability of the objective shooting conditions and the delay of the scene caused by the interference,but also makes the detection accuracy of the remarkably changed image area high.It achieves a stable,efficient and accurate automatic detection of the road illegal structures.

关 键 词:无人机 直方图规定化 SURF 边缘检测 形态学滤波 

分 类 号:TP301[自动化与计算机技术—计算机系统结构]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象