An Integrated Bivariate Function Fitting Method and Its Application in XRII Image Distortion Correction

作者：Yan Shiju; Song Chengli^*

来源：Journal of Medical Imaging and Health Informatics, 2014, 4(3): 433-438.

DOI：10.1166/jmihi.2014.1271

• X-Ray Image Intensifier
• Distortion Correction
• Moving Least Squares
• Multilevel B-Spline
• Function Fitting

The aim of this study is to develop and evaluate a new integrated bivariate function fitting method to correct X-ray image intensifier (XRII) image distortion. The proposed method (iMM) integrates Moving Least Squares method (MLS) and multi-level B-spline approximation method (MBA). MLS is used to generate denser "virtual" data points; MBA is applied to fit the ultimate function based on the generated data points. Experimental results using computer-simulated data and real XRII images showed that the proposed method, iMM, possesses higher fitting accuracy than traditional local and global methods, moreover, iMM is less sensitive to pincushion and sigmoidal distortion than both traditional methods. iMM could be a better choice to correct XRII image distortion in the following conditions: (1) pincushion distortion and sigmoidal distortion exist simultaneously, (2) the original control points are sparse, (3) reusability of the correction function is required.

• 出版日期2014-6
• 单位上海理工大学