The invention discloses a traffic signal lamp fault detection method based on image processing. The method comprises the following steps: firstly, acquiring a real-time video of a traffic signal lamp, performing lamp group segmentation, then judging whether the type of the signal lamp is a red-yellow-green signal lamp, if so, judging a signal lamp fault by adopting a comprehensive discriminator, and outputting a signal lamp fault type; and if a countdown board is a countdown board, preprocessing the countdown board, then dividing the countdown board into seven sections, respectively judging the fault of each section by adopting a comprehensive discriminator, and finally judging the fault type of the whole countdown board by integrating the fault states of the seven sections. The comprehensive discriminator calculates the fault probability P1 according to the on-off state, calculates the fault probability P2 according to the on-off fluctuation, calculates the fault probability P3 according to the SVM classifier, and finally combines the three probabilities to judge the fault state of the signal lamp. The method can be used effectively detect fault types such as long-time turn-off and long-time turn-on of traffic lights, simultaneous turn-on of two lights and abnormal countdown board, and has very high practicability.
本发明公开了一种基于图像处理的交通信号灯故障检测方法。首先获取交通信号灯的实时视频,进行灯组分割,然后判断信号灯的类型是否为红黄绿信号灯,如果是红黄绿信号灯,则采用综合判别器判别信号灯故障,并输出信号灯故障类型;如果是倒计时牌,则先进行预处理,然后分为7段分别采用综合判别器判别各段故障,最后综合上述7段的故障状态判定整个倒计时牌的故障类型。上述综合判别器根据亮灭状态计算故障概率P1,根据亮灭波动计算故障概率P2、根据SVM分类器计算故障概率P3,最后综合三个概率判断信号灯的故障状态。本发明可以有效检测交通信号灯长灭、长亮、双灯同亮、倒计时牌异常等故障类型,具有很强的实用性。
Traffic signal lamp fault detection method based on image processing
一种基于图像处理的交通信号灯故障检测方法
2022-02-11
Patent
Electronic Resource
Chinese
Multifunctional traffic signal lamp fault detection equipment
European Patent Office | 2022
|