针对扭曲的XRII图像给C臂X光机成像系统实时标定带来不利影响的问题,提出了基于校正靶特征和Biharmonic样条曲面插值相结合的XRII图像扭曲校正方法.在分析校正靶自身特征的基础上,使用Otsu算法与形态学相结合的方法提取了校正靶网格线交叉点坐标,然后拟合提取出的网格线交叉点坐标,最后使用Biharmonic样条曲面插值方法校正扭曲的XRII图像.实验结果表明,使用由Otsu算法与形态学相结合的方法提取的校正靶网格线交叉点坐标值与理想校正靶网格线交叉点坐标值的残留误差和标准误差分别为10.489 0±0.074 0和4.263 2±0.015 0,然而使用Biharmonic样条曲面插值校正之后残留误差和标准误差分别降低至(7.121 85±3.354 15) ×10-13和(3.915 1±1.862 7)×10-13.%To solve the problem that distortional XRII(X-ray image intensifier) image brings about adverse effect on real time calibration for C-arm CT (computer tomography) imaging system, a distortion correction method for XRII image based on calibration grid characteristic and Biharmonic spline surface interpolation is proposed. After analysing calibration grid characteristic, a connection method of Otsu algorithm and morphology is proposed to extract the coordinates of intersect marker on calibration grid. Then the extracted coordinates are used to fit a surface. Finally, Biharmonic spline surface interpolation are used to correct the distorted XRII image. Experiment demonstrates that the residual error and the standard error between the extracted intersect marker coordinates on calibration grid using Otsu algorithm and morphology and the intersect marker coordinates on calibration grid of ideal image are 10.489 0 ±0.074 0 and 4.263 2 ±0.015 0, respectively. However, after using Biharmonic spline surface interpolation, the residual error and the standard error are respectively reduced to (7.121 85 ±3.354 15) x10-13and (3.915 1 ±1.862 7) x10-13.
展开▼
