融合仿射与透视变换模型的干涉图像双级配准方法OA

In oblique-incidence laser interferometry system,the geometric distortions arising from high-incidence-angle measurement beams,characteristics of measured helical tooth surfaces,and optical components can compromise registration accuracy.To address this issue,this paper proposes a dual-stage registration method for dental surface interferograms that integrates affine and perspective transformation models.Firstly,the sources of distortion in dental surface interferograms are analyzed.The Shi-Tomasi corner detection algorithm is then employed to extract boundary feature points from the measured tooth surface images.An affine transformation is applied to perform global registration of the simulated tooth surface images with respect to the measured ones.On this basis,a perspective transformation model is fur-ther introduced to achieve finer local registration between the two sets of images.Finally,comparative ex-periments are designed to evaluate the proposed method against three classical registration algorithms.Ex-perimental results demonstrate that the proposed method improves the average distance index between measured and simulated boundary point pairs by approximately 78%,effectively enhancing both registra-tion accuracy and efficiency.