节点文献
利用神经网络自动识别X线头颅影像标记点
Automatic Localization of Landmarks for Cephalometry by BP Neural Network
【摘要】 目的:研究并设计能够自动完成对X线颅颌面影像定点分析的实用程序。方法:采用数学形态学技术对图像增强,识别并提取子图像缩小标志点的标定范围,利用神经网络识别并定位标志点。结果:利用本文所述的方法设计的识别程序很好地完成了对测试X线颅颌面影像的自动定点分析。结论:此法能够高效且准确地对X光颅颌面影像进行自动定点分析。
【Abstract】 Objective:To design an automatic cephalometric program to analyse the cephalogram.Methods:The images were enhanced by morphologic method,and then recognized and segmented into sub-images.The landmarks were recognized by BP neural network.Results:The program accurately identified landmarks of the test images.Conclusion:The method can automatically analyse and identify the cephalograms fast and accurately.
- 【文献出处】 放射学实践 ,Radiologic Practice , 编辑部邮箱 ,2006年10期
- 【分类号】R816.98
- 【下载频次】50