Штучні нейронні мережі в задачах створення 3d моделей на основі зображення об’єкта

Abstract

This proposed work is devoted to image processing in photogrammetry and transformation of 2D images into 3D models using deep neural networks. Existing solutions provide 3D models after the user has played a large number of actions such as filming a video or taking a large number of pictures. The purpose of the work is to minimize the actions taken to obtain a 3D model. To achieve this goal, the use of generative adversarial neural networks is considered.