摘要
This paper presents a dynamic detection method of medium and low speed maglev F-track seam based on machine vision detection technology. A Kalman filter light bar center extraction algorithm based on ridge line prediction is proposed for the bar image of F-track seam detection. And a threshold segmentation feature point extraction method is proposed for the center of the light bar. The static and dynamic precision of detection system are verified by the stability test, static and dynamic precision test. The static detection accuracy of end of F-track seam is less than 0.5mm and dynamic detection accuracy is less than 0.8mm. Then the static precision is less than 0.2mm the dynamic accuracy is less than 0.5mm about the horizontal F-track seam. And the static precision is less than 0.2mm, and the dynamic accuracy is less than 0.5mm about the vertical F-track seam. The detection accuracy of the system meets the requirements of F-track seam precision.
