Research On The Warning Technology Of Vehicle Abnormal Running Based On Android Vehicle System

Vehicle warning system of abnormal running will be an important part in driving assistant systems, it will judge whether the vehicle is traveling in an abnormal state through the analysis of driver information, environmental information, condition information. When the vehicle is potentially in hazardous situation, it will warn the driver. In the vehicle warning system, the information acquisition and characteristics processing of drivers, the environment and condition are very important. The main work in the paper is the research of the warning technology of abnormal running behavior based on the Android vehicle system, and results are as follows:1.An analysis of the reason which may cause abnormal vehicle driving, then tell the characteristics of these acts, and categorized them into three categories:misoperation, illegal driving and vehicle defects;2. Reprogram Camera program according to the Android API, convert the format of the video stream from YUV420SP to RGB, which optimizes the process of call real-time video streaming and improves the real-time of early warning system;3. Model the early warning of vehicle safe driving, following three sub-models:early warning model of illegal driving, misoperation and vehicle defects, there are corresponding identificating algorithm of abnormal behavior under the sub-models;4. Verify the accuracy of detecting algorithm on a PC, and then use JNI technology to port detecting algorithms writed by C++ to the Android platform, to verify the android program’s real-time, stability, and running accuracy.Experiments show that the early warning model of vehicle safe driving may be more effective and reliable to monitor vehicles’ abnormal traveling, help the driver better grasp the state of the vehicle, such as unusual early warning that drivers could take corrective action immediately and promptly, reduce the incidence of accidents.