Research On Vehicle Corner Identification And Relative Positioning Based On Smartphone

When the vehicle is traveling on a curved road, due to drivers’ unfamiliar technical or driving violations or other reasons, accident like sliding, collision and so on, can easily occur. Therefore, traffic safety on curve road is particularly important. Among them, the lane recognition, especially the curve recognition played a significant role in the field of intelligent transportation. Meanwhile, the smart phone as a good carrier for intelligent transportation field can provide a variety of functional applications. The use of smart phones in the field of intelligent transportation for device development has become a new way of thinking and problem solving. This study is based on smart phones and the issues to be studied are curve identification and relative positioning of vehicles.First, we study the selection of sensors to detect curve. In this paper, when vehicle traveling on the road, the actual data are collected, and the data are classified according to the type of curve. Use fitting test which belong the hypothesis testing to finalize the sensor and its variable value which the curve recognition desired. Experiments show that the sensor and the variable value which are filtered out can make the curve recognition algorithms achieve more than 90 percent accuracy rate.Secondly, we study curve recognition algorithm based smart phone sensors. Based on the exploration of sensor which curve recognition needed and the characteristics of the experimental data, this paper presents a curve recognition method which fused the nearest criterion and the chi-square statistic, to explore the impact of data collected by sensors on curve recognition. A large number of experiments show that the algorithm has a good recognition of seven types of road and it’s accuracy rate can achieve more than 90 percent.Finally, this paper studied the vehicle relative position localization algorithm which is based on the smart phone. The algorithm is divided into two scenarios, one is a straight road scene, the other is curve road scene. This paper proves that the relative position localization algorithm of the vehicle in a straight road is not available for the relative position location algorithm of the vehicle in a curve road, and proposed a new vehicle relative position localization algorithm on curve road which fused the curve recognition algorithm and the wifi signal strength values of the phone. Experimental results show that the average accuracy of location can achieve more than 90 percent and does not depend on the weather and the type of vehicle.