Research on Embedded Rail Surface DefectDetection System Based on Multi Core DSP

摘要

Track surface defect detection is a key link in daily inspection of high-speed railway. Combined with the adaptive requirements of modern railway and automatic detection technology. Accurate and rapid detection of rail surface defects has become an important guarantee for the safe and stable operation of trains. Aiming at the problem of fast and real-time detection of rail surface defects in high-speed railway long tunnels, a multi-channel embedded rail surface defect image detection method is proposed. The method adopts the image acquisition of DSP cooperative array, As the front-end component of rail surface defect detection, storage equipment realizes a multi-channel embedded high-speed rail surface defect detection system. Experiments show that on the platform of DSP integrated development software CCS 3.1 (Code Composer Studio), the accuracy of the system for detecting rail surface defects reaches 95.37%. The identification time of rail surface defects is within 4.5 ms, and this method can meet the requirements of the real-time detection system for rail surface defects in long and high-speed railway tunnels in terms of processing speed and accuracy, and can replace the manual inspection of rail surface defects to some extent.