摘要
路面结冰、积雪易造成飞机起降的安全问题,实时监测机场跑道冰雪情况成为防止事故发生的必要手段。文中提出采用机器视觉技术对路面冰雪厚度进行监测,从系统性能设计的角度确保量程和精度要求。针对待测冰雪厚度、边缘形状、待测物个数的随机性以及图像噪声大的特点,采用保边滤波和中值滤波相结合的方式去除噪声,在Canny算子的基础上设计了基于边缘特征的一阶差分法和自适应阈值法来识别有效边缘,多次实验验证表明,两种方法的平均检测误差小于10%。
Ice and snow on the road can easily cause safety problems for aircraft taking off and landing.Real⁃time monitoring of the thickness of ice and snow on airport runways has become a necessary means to prevent accidents.This paper proposes the use of machine vision technology to monitor the thickness of ice and snow.The range and accuracy requirements are ensured from the perspective of system performance design.Aiming at the problem of the thickness,edge shape and number randomness of the object under test,and large image noise,the edge⁃preserving filters and median filter are combined to remove the noise,and based on the Canny operator,this paper designs the first⁃order difference method based on edge features and the adaptive threshold method to identify effective edges.Multiple experiments verify that the average detection error of the two methods is less than 10%.
作者
吴琼
薛宁
刘春秀
WU Qiong;XUE Ning;LIU Chunxiu(State Key Laboratory of Transducer Technology(Northern Base),Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China)
出处
《电子设计工程》
2022年第12期11-17,共7页
Electronic Design Engineering
基金
国家自然科学基金资助项目(62073307,61774157)。
关键词
机器视觉
结冰探测
成像质量
图像去噪
边缘检测
machine vision
icing detection
imaging quality
image denoising
edge detection
