Research On The Detection Algorithm Of Non-safe Driving Behavior Based On Machine Vision

The appearance of mobile phones has brought great convenience to people’s life,and it also has a great impact on the safe driving of cars.Nowadays,non-safety driving behaviors such as making and receiving calls or looking down at mobile phones during driving are becoming more and more common,which has become one of the most important causes of frequent automobile accidents.At present,there are not many studies on the detection of the behavior of drivers using mobile phones during driving.The commonly used detection method is to judge through the acquisition of the communication signal of the mobile phone,but this method can not accurately identify whether the mobile phone signal is from the driver or the passenger.Based on this,this paper proposes a method of detecting unsafe driving behavior based on machine vision.The method firstly collects the driver’s driving behavior data through the camera installed on the windshield of the vehicle,and then uses relevant algorithms to analyze the collected video data.Finally,whether there is any unsafe driving behavior using the mobile phone during the driving process is monitored in real time.Firstly,this paper proposes a face detection algorithm based on the combination of skin color segmentation and AdaBoost algorithm,and proposes a scale-adaptive KCF target tracking algorithm on face tracking,and completes the related contrast verification experiments.Experimental results show that the proposed algorithm has better performance.Secondly,based on the face detection results,the regional positioning of hand image candidate regions for non-safe driving behavior was completed,and a hand image detection method combining YCbCr skin color extraction and skin color threshold was proposed.The method of averaging the results of multi-frame image detection is used to judge whether there is any unsafe driving behavior using the mobile phone during driving.Finally,the reliability and accuracy of non-safe driving behavior detection method based on machine vision are verified through experiments.The research results show that the method of non-safe driving behavior detection based on machine vision can effectively detect the driver’s unsafe driving behavior during driving.And it can be practically applied to the safe driving warning of public transport vehicles such as buses,long-distance passenger cars,and taxis,effectively improving traffic safety and reducing the occurrence of traffic accidents.