Vehicle Accident Detection and Identification Using Image Compression Analysis and RFID Traffic Cone Tracking System Module

This paper proposed an intelligent RFID traffic cone that is applied for vehicle accident detection and identification based on image compressing analysis and RFID detection tracking in an accident clamming system and traffic reporting system. The gain benefit of this paper can reduce the waiting time of an accident clamming process that usually used along time with normally vehicle crash accessing and reporting systems, especially those related to an image processing and insurance techniques. This RFID technique deals with multivehicles, multi lane and multi road even or traffic junction area. It provides an efficiency time management scheme with enough correctly data reporting, in which a dynamic time schedule is worked out in real time for the driver or passengers of each accident situations. The accessing time operation of the RFID traffic cone system emulates the judgment of a traffic policeman on duty or user who may have the PDA nearby the RFID traffic cone that can be connected via by Wireless channels. The image compression present here is used along with the RFID information of each vehicle to get a precise event data picture that is composed of image encoding and decoding algorithms called wavelet transform with Principle Component Analysis (PCA) via Vector Quantization techniques (VQ). The small bit rates for high-speed data transmission with a small space for data storage area are required on wireless transmission channel. Simultaneously, the peak signal to noise ratio (PSNR) has to be maintained. The traffic management system model is constructed for testing this present idea that composed of traffic lights, vehicles transit, vehicles clash and traffic cone with RFID solution system. By applying the proposed technique, performance has been improved which indicated by lower bit rate and better PSNR for an image compression algorithm. The RFID traffic cone mechanisms are work well while the RFID data tags that are recorded of each vehicle are enough correctly and can be sent the data to the traffic information center or an accident clamming center via on local area network (LAN) or wide area network (WAN) simultaneously. For scaling the large number of vehicles in a real situation, the simulation model is used to test this system and reviewed that this proposed technique may be work well. KeywordsRFID Traffic Cone; PCA; VQ; DWT; CLC; CLC+SEC