基于改进光线投影算法的医学图像三维重建

发布时间：2018-05-20 09:52

本文选题：光线投影 + 采样频率　；参考：《计算机应用研究》2017年12期

【摘要】：针对经典光线投影算法在医学图像三维重建中渲染速度慢的问题进行了研究,提出一种改进的光线投影算法。首先在每条经过三维数据场的射线上,根据观察者视点与图像远近的不同,使用曲面法线切点处的偏导数与视点和切点之间的距离的比例关系来调整采样频率,改变经典算法对所有射线上的采样频率都一视同仁的做法,实现当物体距离视点较近时,采样频率较高,模型表示较为细致,反之则采样频率较低,模型表示也较粗糙;其次在图像合成算子的计算上,设定对未到达屏幕的射线不参与不透明度和颜色值的合成计算。临床数据实验证明,改进算法不仅能够有效提高图像渲染速度,而且还避免了伪影的产生,明显改善了图像的质量。
[Abstract]:In this paper, the problem of slow rendering speed of classical ray projection algorithm in 3D reconstruction of medical image is studied, and an improved ray projection algorithm is proposed. First, on each ray passing through the 3D data field, the sampling frequency is adjusted by using the proportional relation between the partial derivative at the tangent point of the normal curve of the curved surface and the distance between the viewpoint and the tangent point according to the distance between the observer's viewpoint and the image. By changing the classical algorithm, the sampling frequency of all rays is treated equally. When the object is close to the viewpoint, the sampling frequency is higher and the model representation is more detailed. Otherwise, the sampling frequency is lower and the model representation is rough. Secondly, in the calculation of the image synthesis operator, the calculation of the opacity and the color value of the rays not arriving at the screen is set. The clinical data show that the improved algorithm can not only improve the image rendering speed, but also avoid the artifacts and improve the image quality.
【作者单位】：上海大学计算机工程与科学学院;上海交通大学医学院附属瑞金医院计算机中心;
【基金】：上海市科技创新行动计划资助项目(16511101200) 数字化医疗医院示范(2012AA02A612)
【分类号】：TP391.41

【相似文献】