Novel method for rail wear inspection based on the sparse iterative closest point method

As trains become progressively faster, it is becoming imperative to automatically and precisely inspect the rail profile of high-speed railways to ensure their safety and reliability. To realize this, a new method based on the sparse iterative closest point method is proposed in this study. Moreover, the noncontact method is mainly used for convenience and practicality. First, a line laser-based measurement system is constructed, and the position of the line laser is calculated to ensure that both the top and sides of the rail are in range of the line laser. Then, the measured data of the rail profile are separated into a baseline part and worn part. The baseline is involved in registering the measured data and reference profile by the sparse iterative closest point method. The worn part is then transformed by the same matrix of the baseline part. Finally, the Hausdorff distance is introduced to measure the distance between the wear model and reference model. The experimental results demonstrate the effectiveness and efficiency of the proposed method.