Font Size: a A A

Research On The Standardization Of Vehicle Speed Calculation System Based On GB/T 33195-2016 Standards

Posted on:2022-04-17Degree:MasterType:Thesis
Country:ChinaCandidate:T Y FangFull Text:PDF
GTID:2492306569450014Subject:Vehicle Engineering
Abstract/Summary:PDF Full Text Request
China officially promulgated and implemented the standard "Vehicle Speed Identification in Road Traffic Accidents"(GB/T 33195-2016)in 2017 to guide the calculation of vehicle speed in road traffic accidents.With the popularity of surveillance video and the development of simulation reproduction software,appraisers are more interested in using efficient,reliable and convenient computer-aided systems for speed calculation.Based on the GB/T 33195-2016 standard,Professor Wei Lang’s team at Chang’an University developed the Truthview system for speed calculation from video footage,and the Crashview system for speed calculation from simulation reproduction software,and verified the reliability of the system.However,in the process of practical application,there are problems such as insufficient adaptability to different computer monitor screens,inability to achieve local scaling of feature points of objects with inappropriate morphological sizes and simple automatic recognition of feature points,etc.The Crashview system has problems such as mismatch between accident scene survey information and the parameters required for the system to perform vehicle speed calculation,and insufficient orientation of the accident scene survey process.These problems affect the practicality,convenience and reliability of the system,and also reduce the efficiency and social credibility of front-line appraisers and law enforcement officers.Therefore,based on the GB/T 33195-2016 standard,it is of great importance to develop a standardised speed calculation system for frontline accident investigation and appraisal personnel based on the Truthview and Crashview systems as the core of the above-mentioned problems,which will complement and improve their functions respectively.The resolution adaptation model is established by exploring the correspondence between different computer monitor resolutions and image resolutions when the image is exactly full on the screen,in response to the lack of computer monitor adaptability of the Truthview system.According to the calculation principle of Truthview system,the resolution of the image after adapting to the screen size is read automatically and passed to the coordinate system positioning module as a parameter value,always fixing the coordinate origin at the lower left corner of the image to ensure the correctness of the subsequent speed calculation;for a large number of sub-frame images,the resolution batch modification module is designed to improve the efficiency of use,reduce memory usage and ensure the stability of the system processing data.It also ensures the stability of the system in processing data.To address the problem that the Truthview system cannot achieve local scaling,the local scaling function of the image is implemented based on the second-order Newtonian interpolation principle.It also explores the mapping relationship between the coordinate system of the Truthview system and the coordinate system of the local scaling module,and establishes a coordinate system transformation model to ensure that the coordinate values of feature points in the Truthview system remain unchanged before and after scaling.Based on the improved hough transform algorithm,a wheel centre feature point recognition model is established for the wheel centre recognition problem.The wheel centre recognition success rate is 88.9% under the condition of vertical wheel side shot,and the error in the horizontal coordinate of the wheel centre point is basically less than 0.5% and the error in the vertical coordinate is less than 1%.This module automatically terminates the calculation when the rate of change of the coordinate value of the wheel centre point in one direction in three consecutive frames exceeds 50%,avoiding the situation that the vehicle speed is calculated incorrectly without knowledge.In view of the mismatch between the accident scene survey information and the required parameters of Crashview system,through the study of the positioning method of the accident scene elements and the positioning method of Crashview system,an automatic coordinate conversion model was established for the straight(curved)east-west road,straight(curved)north-south road,cross-shaped road,combined straight-curved east-west road,combined straight-curved north-south road,T-junction,etc.based on the X-Y coordinate system of Crashview system.At the same time,for the problem that the parameters required by the Crashview system cannot be directly located in the accident scene survey,the internal coordinate conversion algorithm model of four-wheeled vehicles,three-wheeled vehicles and two-wheeled vehicles is established,which greatly reduces the amount of parameter input and the difficulty of obtaining.This greatly reduces the amount of parameter input and the difficulty of acquisition,and further improves the practicality of the system.To address the problem of inadequate orientation of the accident scene survey process,a set of standardised accident scene survey processes for the Crashview system was designed by studying the accident scene survey standards and the common accident element positioning methods used by frontline surveyors,combined with the Crashview system coordinate system,positioning methods and the parameters required for calculation.By following the process step by step,the surveyor can seamlessly transfer the accident scene location information to the Crashview system,further improving the efficiency of the Crashview system.Based on the key technologies implemented to solve the above problems,the Truthview and Crashview systems are benchmarked,and the Visual C++ 6.0 programming platform is used to develop an image processing and feature point recognition system based on the Truthview system for the method of calculating vehicle speed using video images,and a simulation reproduction software for the method of calculating vehicle speed using a coordinate reconstruction system for road traffic accident scene based on Crashview system.The complete speed identification process from video image processing or standardized survey of the accident scene to speed calculation and final identification report output under different identification material conditions is realized,which is of great significance to realize the standardization,normalized and regularized of speed identification of road traffic accidents in China.
Keywords/Search Tags:GB/T 33195-2016, Truthview, Crashview, Vehicle Speed Calculation, Image Processing Recognition, Coordinate Reconstruction, Accident Scene Investigation, System Dvelopment
PDF Full Text Request
Related items