作者: Madjid Maidi , Yassine Lehiani , Marius Preda
DOI: 10.1109/ICIP40778.2020.9191362
关键词: Pose 、 Artificial intelligence 、 Augmented reality 、 Computer vision 、 Graphics 、 Relevance (information retrieval) 、 Transformation (function) 、 Visual perception 、 Computer science 、 Cognitive neuroscience of visual object recognition
摘要: The aim of this work is to present an open solution for building Augmented Reality (AR) system without using any existing SDK. proposed approach relies upon 2D planar object recognition mobile real-time tracking applications. transformation relating the world and camera coordinate systems determined pose estimation. Once projective transform 3D features computed, a virtual graphic registered on image. Many tests have been performed show efficiency prove its relevance in terms accuracy time computation. final application enabled markerless images augmented with models enrich visual perception user.