基于改进光线投影算法的医学图像三维重建

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。

Three Dimensional Reconstruction of Medical Images Based on Improved Ray Projection Algorithm

In order to improve the rendering speed of the classical ray-casting algorithm in 3D reconstruction of medical image, this study proposes an improved algorithm of ray projection. On one hand, according to the observer viewpoint and the distance between different images, this study used surface normal line and focused on the point of tangency of the partial derivative with viewpoint and point distance ratio to adjust the sampling frequency which through the three-dimensional data field in each ray. The observer can get more detailed model representation when the object distance is close to meet the viewpoint. On the other hand, some rays which are not reaching the screen were not involved in the synthesis of ray opacity and color value calculation. The experimental results show that the improved algorithm in this study not only can effectively improve the image rendering speed, but also avoid the generation of artifacts, which can improve the quality of the image to a certain extent.