Automatic Image Calibration And Its Application In Steel Rail Abrasion Measurement System

Visual measurement has been widely used in industrial inspection for its non-contact, high efficiency and automatic. This dissertation is based on the research of the optical structure method of visual measurement system. The calibration technique in the single camera vision measurement system has been discussed.Calibration is one of the key technologies of vision measurement system which determines the accuracy of the measurement system. The calibration includes the determination of the sensor’s structure parameters and the intrinsic parameters and distortion coefficients of the camera. In this dissertation the distortion of the imaging system has been discussed detailed, an imaging model which considers a variety distortion effect was put up. The proposed method effectively solves the consistency problem of the system’s calibration status and working status. The calibration process is simple and the accuracy is properly.The research background of this dissertation is the calibration of the image in the rail wear image measurement system. A variety calibration methods were analyzed and proved that a single image can be used to extract its world coordinates to determine its geometric dimensions in the measure plane. An implicit parameter calibration method is proposed to direct reversing the image to the measure plane coordinate system. We made a dedicated calibration plate, and developed a calibrate software based on MFC which uses the principle of least squares to reduce the calibration error, and verified calibration results. The rail wear measurement system was calibrated with this method, and archived the desired results through laboratory and field measuring, which further proofed that calibration is feasible and effective. The GHC-1rail head wear laser image measurement and analysis device passed the identification of scientific and technological achievement.