1. Application of Spatial Registration Algorithm Based on ORB Features in Augmented Reality
- Author
-
Xuzhi Wang and Xiangjie Li
- Subjects
0209 industrial biotechnology ,Computer science ,Computation ,Decision tree ,Process (computing) ,02 engineering and technology ,RANSAC ,020901 industrial engineering & automation ,Feature (computer vision) ,0202 electrical engineering, electronic engineering, information engineering ,020201 artificial intelligence & image processing ,Augmented reality ,Point (geometry) ,Algorithm ,Orb (optics) - Abstract
In consideration of problems with traditional augmented reality, including inaccuracy, untimeliness and instability of 3D spatial registration results, this paper proposes an improved algorithm based on ORB (Oriented FAST and Rotated BRIEF) features, which improves recursive adjustment procedures for decision trees during feature point extraction and makes up deficiencies of conventional ORB algorithms such as high computation load and inadequate feature point extraction. In terms of 3D spatial registration, this paper establishes an augmented reality system by RANSAC optimization algorithm. The experimental results suggest that it costs relatively short time to process complicated natural images with the algorithm proposed in this paper, which is highly timely and robust for processing scene images.
- Published
- 2018
- Full Text
- View/download PDF