论文部分内容阅读
目的 :实现头颅侧位片标志点的自动识别 ,减少人工定点带来的误差。方法 :确定标准模板并确定标准图像上标志点的位置 ,利用仿射变换将这些点变换到被测图像上作为初始位置 ,以缩小搜索范围 ;利用对标准模板的形变实现不同标志点的识别。结果 :实现软硬组织轮廓的提取和一些孤立点的识别 ,对依附于轮廓的点的精度较高。结论 :该方法能较好的识别头颅侧位片中常用的测量点。
OBJECTIVE: To realize the automatic recognition of the marked points of the skull flaps and to reduce the error caused by the artificial pointing. Methods: Determine the standard template and determine the location of the mark on the standard image. Use the affine transformation to transform these points to the measured image as the initial position to reduce the search range. Use the deformation of the standard template to realize the recognition of different mark points. Results: The extraction of soft and hard tissue contours and the identification of some isolated points resulted in higher accuracy of the points attached to the contour. Conclusion: This method can identify the commonly used measurement points in cephalometric films.