Self-calibration method for spiral CT geometric parameters based on data redundancy conditions.

The geometric parameters of the CT system significantly influence the quality of reconstructed images and the accuracy of dimensional measurements. To achieve precise calibration of these parameters, this study develops a self-calibration method for spiral CT geometry based on data redundancy conditions. Theoretical analysis is conducted to identify the geometric relationships that satisfy data redundancy under spiral scanning trajectories. Using simulation data, we investigate the effects of three key geometric parameters-the source-to-sample distance, the horizontal offset of the detector, and the rotation angle in detector plane-on data redundancy. Additionally, we examine how varying noise levels impact the accuracy of geometric parameter estimation. Finally, the proposed method is validated through spiral CT scanning experiments with real samples. Results show that the calculation accuracy for the detector's horizontal offset is better than 1 pixel, the rotation angle in detector plane accuracy exceeds 0.1°, and the source-to-sample distance accuracy surpasses 0.42%. These findings confirm the effectiveness of the proposed method in practical applications.