The manual measurement of high-speed train axles presents limitations in efficiency and consistency, becoming a bottleneck in high-volume manufacturing. This study introduces an automated optical measurement system as a solution. The process involves system calibration, point cloud denoising, and diameter calculation via an improved bat algorithm (IBA) for optimal circle fitting. Our experiments on real axles show that the system achieves a precision of ±0.02 mm. This performance demonstrates that the proposed system is a viable and accurate tool for automated axle diameter measurement in industrial settings.