Abstract:Unmanned Aerial Vehicle (UAV) remote sensing image registration is the key step of remote sensing image stitching, image fusion and multi-frames image super-resolution. Its speed and accuracy determines the effect of remote sensing image applications, such as object detection, environment monitoring. To meet the speed and accuracy requirements of UAV remote sensing image registration, an improved Oriented FAST and Rotated BRIEF- Random sample consensus (ORB-RANSAC) algorithm is proposed. Firstly, images to be registered are divided into non-overlapping sub-images, and then a simplified image pyramid is constructed for these sub-images to get scale invariance. Secondly, the traditional FAST corner detection algorithm is improved by setting the adaptive corner detection threshold, and more feature points are detected. Meanwhile, the traditional quadtree algorithm is improved to remove redundant feature points and keep the remaining high-quality feature points. Thirdly, feature points coarse matching is done by bidirectional matching combined with cosine similarity method. Finally, the improved RANSAC algorithm is used for feature points fine matching to eliminate mismatches and calculate the transformation matrix. Experiment results show that, comparing with the traditional ORB algorithm, the number of feature points detected is significantly increased and its distribution is more uniform, and the correct matching rate increases by 58.10% in the case of image scale changing. Comparing with the state-of-the-art UAV remote sensing image registration algorithm, the correct matching rate and mutual information of our method are increased by 0.68% and 1.91% respectively, matching time and root mean square error are reduced by 3.89% and 11.2% respectively.

Improved Augmented Reality Registration Method Based on VSLAM

Research on marker tracking method for augmented reality.

VSLAM method based on object detection in dynamic environments

Augmented reality registration from nature features based on planar color distribution

A SLAM-Based Mobile Augmented Reality Tracking Registration Algorithm

A Survey of Monocular Simultaneous Localization and Mapping

Survey and Evaluation of Monocular Visual-Inertial SLAM Algorithms for Augmented Reality

A 3D Tracking and Registration Method Based on Point Cloud and Visual Features for Augmented Reality Aided Assembly System

Local Optimized and Scalable Frame-to-model SLAM

Unmanned aerial vehicle remote sensing image registration based on an improved oriented FAST and rotated BRIEF- random sample consensus algorithm

A fast object registration method for augmented reality assembly with simultaneous determination of multiple 2D-3D correspondences

A Tracking Registration Method for Augmented Reality Based on Multi-modal Template Matching and Point Clouds

Robust Keyframe-Based Monocular SLAM for Augmented Reality

AR Based on ORB Feature and KLT Tracking

A semantic visual SLAM based on improved mask R-CNN in dynamic environment

A practical marker-less image registration method for augmented reality oral and maxillofacial surgery

An improved SLAM based on RK-VIF: Vision and inertial information fusion via Runge-Kutta method

VSLAM based on deep learning in low-textured scenes

Fast Automatic Registration of UAV Images via Bidirectional Matching

Increasing SLAM Pose Accuracy by Ground-to-Satellite Image Registration