上海科技大学知识管理系统

本文描述了一种用于构建三维图像的计算机实现的方法。计算系统可以获得对象的多个稀疏视图正弦图。计算系统可以对机器学习模型进行编码以表示对象的连续密度场，从而将体素连续密度场的空间坐标映射到强度值。机器学习模型可以包括被配置为通过多分辨率散列编码技术将连续密度场的体素映射到更高维空间的编码模块，以及被配置为根据连续密度场的体素的空间坐标输出强度值的神经辐射模块。计算系统可以使用多个稀疏视图正弦图来训练机器学习模型。计算系统可以确定要构建的对象的三维图像的体素的空间坐标。计算系统可以将三维图像的体素的空间坐标输入到经训练的机器学习模型中。计算系统可以基于由经训练的机器学习模型生成的密集视图正弦图来构建三维图像。

Described herein is a computer-implemented method for construction of three-dimensional images. A computing system can obtain a plurality of sparse-view sinograms of an object. The computing system can encode a machine learning model to represent a continuous density field of the object to map spatial coordinates of voxels continuous density field to an intensity values. The machine learning model can comprise an encoding module configured to map voxels of the continuous density field to a higher dimensional space through a multiresolution hash encoding technique, and a neural radiance module configured to output intensity values in accordance with spatial coordinates of the voxels of the continuous density field. The computing system can train the machine learning model using the plurality of sparse-view sinograms. The computing system can determine the spatial coordinates of the voxels of the three-dimensional image of the object to be constructed. The computing system can input the spatial coordinates of the voxels of the three dimensional image into the trained machine learning model. The computing system can construct the three-dimensional image based on a dense-view sinogram generated by the trained machine learning model.