Skip to content

🚀 Deep learning includes superpoint-superglue(C++, TensorRT), and traditional algorithms include zkaze, surf, ORB, etc.

Notifications You must be signed in to change notification settings

Broad-sky/feature-detection-matching-algorithm

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

feature detection and matching algorithm models

Introduction

🚀🚀This warehouse mainly uses C++ to compare traditional image feature detection and matching, and deep learning feature detection and matching algorithm models. Deep learning includes superpoint-superglue, and traditional algorithms include AKAZE, SURF, ORB, etc.

  1. akaze feature point detection and matching display.

akaze-image

  1. superpoint-superpoint feature point detection and matching display.

akaze-video

Dependencies

All operating environments, please strictly follow the given configuration,the configuration is as follows:

OpenCV >= 3.4

CUDA >=10.2

CUDNN>=8.02

TensorRT>=7.2.3

How to Run

  1. build.
cd feature-detection-matching-algorithm/
mkdir build
cd build
cmake ..
make
  1. run camera.

deep learning algorithms.

./IR --deeplearning --camera 0

traditional algorithms.

./IR --traditional  --camera 0
  1. run image-pair.

deep learning algorithms.

./IR --deeplearning --image-pair xx01.jpg xx02.jpg

traditional algorithms.

./IR --traditional  --image-pair xx01.jpg xx02.jpg

Models TRT

https://pan.baidu.com/s/1CoK_KuC42BFD-mtO-BBhHg Code:cb7x

TODO

  • Optimizing post-processing using custom TensorRT layer or Cublass.
  • Model conversion script.
  • support for FP16/INT8.

Discussion

WeChat ID: sigma1573

Welcome to add WeChat(note: unit + name) and join the group discussion!

For more details, please refer to zhihu: https://zhuanlan.zhihu.com/p/518877309

SuperPoint

Superpoint pretrained models are from magicleap/SuperPointPretrainedNetwork.

SuperGlue

SuperGlue pretrained models are from magicleap/SuperGluePretrainedNetwork.

Reference

@inproceedings{sarlin20superglue,
  author    = {Paul-Edouard Sarlin and
               Daniel DeTone and
               Tomasz Malisiewicz and
               Andrew Rabinovich},
  title     = {{SuperGlue}: Learning Feature Matching with Graph Neural Networks},
  booktitle = {CVPR},
  year      = {2020},
  url       = {https://arxiv.org/abs/1911.11763}
}

About

🚀 Deep learning includes superpoint-superglue(C++, TensorRT), and traditional algorithms include zkaze, surf, ORB, etc.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published