Feature_Tracker

KO | EN

gitlite — search

#ai #rust #docker #macos #nodejs #java #windows #awesome #cli #kubernetes #awesome-list #security #openai #mcp

Feature_Tracker

(★ 15)

A simple optical flow feature tracker, direct method tracker and descriptor matcher.

Feature_Tracker 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Feature tracker
Intergration of some simple sparse feature track method.

# Components
- [x] Optical flow tracker
  - [x] Basic klt
    - [x] Direct
    - [x] Inverse
    - [x] Fast
  - [x] Affine klt
    - [x] Direct
    - [x] Inverse
    - [x] Fast
  - [x] Lssd klt
    - [x] Direct
    - [x] Inverse
    - [x] Fast
- [ ] Dense optical flow tracker
  - [x] RAFT
  - [ ] Farneback
- [x] Direct method tracker
  - [x] Direct
  - [x] Inverse
  - [ ] Fast
- [x] Descripter matcher
  - [x] Nearby matching
  - [x] Force matching
- [ ] NN feature matcher
  - [ ] Lightglue - superpoint

# Dependence (C++)

### 项目仓库依赖（源码在 repo 中）
- Slam_Utility
- Feature_Detector
- Visualizor2D

### 第三方仓库依赖
- Eigen3 (>= 3.3.7)（`sudo apt install libeigen3-dev`）
- onnxruntime（NN feature match 需要，需下载后放至指定目录，官网：https://github.com/microsoft/onnxruntime）

# Dependence (python)

### 第三方仓库依赖
- pytorch（`pip install torch torchvision`）

# Compile and Run
- 第三方仓库的话需要自行 apt-get install 安装
- 拉取 Dependence 中的源码，在当前 repo 中创建 build 文件夹，执行标准 cmake 过程即可
```bash
mkdir build
cmake ..
make -j
```
- 编译成功的可执行文件就在 build 中，具体有哪些可执行文件可参考 run.sh 中的列举。可以直接运行 run.sh 来依次执行所有可执行文件

```bash
sh run.sh
```

# Tips
- 欢迎一起交流学习，不同意商用；

Feature_Tracker

Feature_Tracker 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Feature_Tracker

Feature_Tracker 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README