Visual Model Based C-Arm System Calibration and Image Correction

Abstract

In traditional methods, the C-Arm system calibration was performed based on the result of C-Arm image correction. When correcting images, high-order polynomial was used to fit relating distorted pixels to ideal image pixel. However accurate ideal image could not be established easily, that heavily restricted the precision of correction. In this paper, a novel method is presented based on visual model which does not depend on ideal image and integrates correction and calibration as one process. In this method, the Tsai's calibration method is used to calculate the internal and external parameters of C-Arm system with markers in the center region of the image. Then, nonlinear optimization is applied to get more precise C-Arm imaging model with all markers. Finally, the distortion image is corrected using the distortion parameters of internal parameters. To verify this method, known distances between points are reconstructed and the experimental results show that the maximum error is less than 1mm.