With the development of Computer Graphics technology as well as deep learning technology, it has become possible to convert 2D images into 3D models using computer algorithms. In the past few years, many implementations of 3D reconstruction techniques have been proposed, some of which have been widely used in areas such as film and television entertainment, intelligent homes, heritage reconstruction, and autonomous driving and etc.
This project aimed to develop a plugin that utilizes a 2.5D reconstruction method based on monocular camera. The purpose of this plugin was to integrate this algorithm seamlessly into DCC tools such as Maya and Unreal Engine, thereby enhancing convenience for users interested in employing this technique. Furthermore, the integration of the plugin into the production pipeline of companies was intended to facilitate the embedding of this algorithm and streamline the workflow within their respective pipelines.
Result (video)
