3D reconstruction method, apparatus, and augment reality AR device

A 3D reconstruction method, apparatus, and augment reality AR device. The method includes: using the AR device to obtain the image information of the object in the current environment, and taking theimage information of each angle of the object as sample data, using neural network algorithm to train the model, and obtaining the 3D reconstruction model of the object. In this way, after the user wear the AR device, the image information of the scanned object is directly processed as the input of the three-dimensional reconstruction model, fast and accurate 3D mesh data can improve the efficiency and accuracy of 3D object model generation, and solve the problem that 3D object model edge jagged or not smooth enough in the traditional 3D reconstruction method, which leads to the difference between 3D object model and actual object. 本申请提供了种三维重建方法、装置及AR设备，利用AR设备获取当前环境中对象的图像信息，并将该对象各角度的图像信息作为样本数据，利用神经网络算法进行模型训练，得到该对象的三维重建模型，这样，在用户穿戴该AR设备后，直接将扫描得到的对象的图像信息作为三维重建模型的输入进行处理，快速且准确得到对象的三维网格数据，提高了生成三维对象模型的效率及精准度，解决了传统三维重建方式得