The Design Of Vehicle Dimensions Measuring System Based On Edge Contour

Machine vision technology has strong stability,widely applicable function,super speed and non-contact characteristics,it is a cross subject in many fields.Now,the technology more extensive apply to vehicles-Dimensions measurement,reverse engineering,mechanical parts quality detection and feature shape recognition.With the growth of the new era of artificial intelligence and hightech,which a sharp increase the number of vehicles has been seriously affected on traffic.To meet demand of intelligent transportation increase the legal supervision of the modified vehicles and guarantee road safety.In this paper,aiming at the intelligent traffic problem of vehicle overload and overrun,the research on the vehicle size measurement is carried out.Focus on the design and structure of software algorithm of measurement system.The non contact measurement and extraction method of the dimension parameters are studied.Main contents of this paper as follows.The methods of camera calibration are studied after analyzing and comparing some common methods.The feasibility of the proposed Open CV camera model calibration method is verified by experiments,and the specific parameters of some vehicles are calibrated.Second,the background subtraction method is used to realize the positioning of the vehicle image with complex background.According to canny image edge detection algorithm of fuzzy edge and the missing problem in machine vision image processing,put forward image block theory and OTSU threshold method to amend the original Canny edge detection algorithm deficiencies.The practical application of the feature extraction algorithm in the vehiclegeometry size measurement system and based on Hough transform to detect the straight line,circle and the Harris theory model is studied in detail.Combined feature matching and region matching technology to achieve stereo matching and complete automatic measurement.The design approaches of vehicle measurement system is studied,developed a friendly PC interface by analyzing system functions.