Data Analysis To The Traffic Checkpoint Based On Cloud Computing

With the improvement of people’s living standard, the number of private cars and many traffic problems such as hit-and-run increase as well, which might lead to serious security risks. The phenominen of vehicles with clone plate is one of them. Due to the activity region of vehicles with clone plate is often far apart and the analysis methods require a combination of vehicles position, it can bring more advantages to detect the vehicles with clone plate by using wider geographical and larger time span traffic data. Howerer, the typical methods for detecting vehicles with clone plate generally run on a single machine, whose limited storage and computing ability restricted the use of mass data to detect vehicles with clone plate to a certain extent. Thus, it is important and significant to find an effective and fast approach to detect vehicles with clone plate in the massive data.According to the principle of unreachable with maximum speed, the thesis presents a distributed detection algorithm based on MapReduce programming model to detect vehicles with clone plate. Meanwhile, we implement the algorithm on both Hadoop and Spark platforms respectively. Compared with typical approaches, the experimental results show that our algorithm is faster and more accurate to detect vehicles with clone plate. Further, combined the detection algorithm and temporal logic abnormal track splitting algorithm, we design and achieve a smart traffic monitor data analysis system based on cloud computing. The system provides users the whole service from data collection to track display of vehicles with clone plate. Compared with other systems, our system not only can run faster than the typical systems, but also can find more vehicles with clone plate which the typical methods may not detect.This paper implements the detection and analysis of vehicles with clone plate in the massive data based on distributed technology. This is not only an innovation of typical methods for detecting vehicles with clone plate, but also an exploration and practice on distributed computing. It has practical significance for big data analysis and its application in the field of transportation.