Top AI Repos — open-source AI, indexed and scored

Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.

jiafeng5513/Evision

计算机视觉实践和探索/Practice and explorations in computer vision.

52No licenceC++

View on GitHub

Stars: 438
Forks: 154
Watchers: 16
Open issues: 5
Contributors: —

Star history, last 90 days— over 2 snapshots

2026-07-31 · 4382026-08-06 · 438

Star history, last 90 days
Date	Stars
2026-07-31	438
2026-08-06	438

Today

— stars today

This week

— stars this week

This month

— stars this month

Momentum

0.0

growth rate 0.00%/day

From the README

<div align=center><img width="100" height="100" src="./src/EvisionSandbox/resource/Evision.ico"/></div>

<div align=center>Evision 双目视觉系统</div>
<div align=center><a href="https://www.bilibili.com/video/av46024738">演示视频</a></div>
<div align=center></div>
<div align=center>如果您觉得有帮助,请为该项目点star.以便于及时收到最新更新.</div>

## 重要提示:Evision 已迁移至 Qt6 + vcpkg 工具链

> **复活与现代化进行中**:本项目正在借助 AI 工具进行阶段性复活与现代化重构。阶段 0(仓库清理)与阶段 1(让 master 重新可编译、CMake 现代化、内存泄漏修复)已完成;内嵌 Darknet 目标检测模块已整体移除(归档于 `archive/darknet-fork` 分支)。后续阶段将引入深度学习立体匹配等 AI 能力。详见下方"路线图"。

Introduction:
=========
1. 双目系统的标定,畸变校正,视差,三维重建,距离测量等.<br>
2. ELAS,ADCensus视差算法.<br>
3. 关于双目的中文资料重复度太高,希望各位后来者能够吸取前人精华,摒弃前人的糟粕,多多自行探索,不要抄来抄去.<br>
4. 关于程序的BUG,以及其他困惑,请使用issues,或联系邮件[email protected].<br>
5. Evision使用 cmake-vcpkg 工具链进行编译,开发者使用 Visual Studio 2019+,Qt6,CMake 3.20+.<br>

资助作者
========
<div align=center><img width="500" height="339" src="./doc/pay.png"/></div>


目录
========
- [Introduction:](#introduction)
- [资助作者](#资助作者)
- [目录](#目录)
      - [1.Dependencies](#1dependencies)
      - [2.Directory_specification](#2directory_specification)
      - [3.build](#3build)
      - [4.模块说明](#4模块说明)
      - [5.双目设备](#5双目设备)
      - [6.参考文献](#6参考文献)
      - [7.数据下载](#7数据下载)
      - [8.路线图](#8路线图)

#### 1.路径说明
1. `data`文件夹存储测试用例(大体积数据需从 Release 下载,见下文).<br>
2. `doc`文件夹中存储的是文档和文档中所用的图片等资源,提供了本项目使用的一些算法相关的论文.<br>
3. `scripts`文件夹存储编译辅助脚本.<br>
4. `src`文件夹中存放代码文件.<br>
* ~~`legacy`~~ 和 ~~`install`~~ 目录已在新一轮清理中移除(旧 C# 版本与编译产物归档于 `archive/legacy-pre-cleanup` 分支).<br>
* ~~`package`~~ 目录已移除(pthread windows 版改由 vcpkg 管理).<br>

#### 2.编译基础组件
1. 安装 Visual Studio 2019 或更高版本 + 使用C++的桌面开发工作负载 + 英文语言包.([visual studio Installer](https://visualstudio.microsoft.com/zh-hans/vs/))<br>
2. 安装 CMake 3.20+.([Download CMake](https://cmake.org/download/))<br>
3. 安装 vcpkg.([GitHub For vcpkg](https://github.com/Microsoft/vcpkg))<br>
4. `vcpkg install opencv:x64-windows`<br>
5. `vcpkg install boost:x64-windows`<br>
6. `vcpkg install pcl:x64-windows`<br>
7. `vcpkg install qt6:x64-windows`<br>
8. `mkdir sln,cd sln`<br>
9. `cmake ..\src\ -DCMAKE_TOOLCHAIN_FILE="${vcpkg_root}/scripts/buildsystems/vcpkg.cmake" -G"Visual Studio 16 2019" -A x64 -DPointCloudViewer=ON`<br>
10. 使用 Visual Studio 打开 `Evision.sln`.<br>
11. 右键 `_CMakeTargets/ALL_BUILD`,生成.<br>
12. 右键 `EvisionSandBox`,设为启动项目.<br>
13. 点击"本地Windows调试器",启动程序.<br>
* 如果vcpkg下载包的速度过慢,可以上代理(以PowerShell为例)`$env:HTTP_PROXY="http://127.0.0.1:1080"`

#### 3.模块说明   
|      |模块名|功能|UI|输出目标|状态|
|:----:|:----:|:----:|:----:|:----:|:----:|
| 1|EvisionADCensus           |ADCensus视差算法  | ×    |动态链接库|✅|
| 2|EvisionElas               |Elas视差算法      | ×    |动态链接库|✅|
| 3|EvisionPnP                |PnP              | ×    |动态链接库|✅|
| 4|EvisionMonocularCalib     |单目标定          | √    |动态链接库|✅|
| 5|EvisionPolyTracker        |单目几何体追踪     | √    |动态链接库|✅|
| 6|EvisionCalibrate          |双目标定          | √    |动态链接库|✅|
| 7|EvisionDisparity          |视差(立体匹配)    |√   |动态链接库|✅|
| 8|EvisionTrace              |交互式测量        | √   |动态链接库|✅|
| 9|EvisionUndistortion       |畸变校正          | √   |动态链接库|✅|  
|10|EvisionCamera             |单目和双目相机功能 | √    |动态链接库|✅|   
|11|EvisionCloudViewer        |三维点云查看      | √    |动态链接库|✅|
|12|EvisionParamBridge        |外部参数传递      | √   |动态链接库|✅|  
|13|EvisionUtils              |通用工具类        | ×   |动态链接库|✅|    
|14|EvisionSandbox            |主程序UI          | √    |可执行程序|✅|

> **EvisionCamera 模块**:已迁移至 Qt6 Multimedia API(`QCameraDevice`/`QImageCapture`/`QMediaCaptureSession`),使用 `QVideoWidget` 替代已移除的 `QCameraViewfinder`。编解码器选择 UI 保留为占位(Qt6 `QImageCapture` 不再暴露编解码器 API)。

#### 4.双目设备
1. 需要注意的是,视差效果,点云效果和精度和设备关系非常大,图片的分辨率越高,光照条件越好,畸变越小,一致性越好,最终效果也就越好.此外,两个相机的距离(基线长度)会影响系统的有效范围,一般来讲,基线越长的双目系统越容易获取远处目标的视差,延长基线能够一定程度上(因为有效距离和相机的焦段也有关)将系统的有效范围拉得更远,但同时,基线越长,盲区(距离相机过近的目标不会同时出现在两个视野中)也越大.
2. 推荐的双目系统:
   1. ZED/RealSense/MYNTEYE小觅相机.这是成熟(昂贵)的商业产品,出厂带有高精度的标定数据和功能强大的SDK,而且还带有IMU,IR主动光学等辅助设备,适合做SLAM,笔者认为购买这类相机是最节约时间成本的方法.<br>
   ![image](./doc/cameras.png)
   1. 双目开发板.淘宝上有很多这类产品,价格比ZED那类便宜很多,同时他们

Excerpt of 7,920 characters

Read on GitHub

Top contributors

No contributor data yet — it is fetched during the contributor sync pass.

Quality score

Would you bet a product on this? Bounded 0–100 and slow moving.

52/10052

Maintenance: 90%; How recently the default branch was pushed to
Releases: 10%; Recency and cadence of tagged releases
Community: 20%; Contributor breadth and bus factor
Issue backlog: 85%; Open issues measured against the size of the audience
Documentation: 83%; README depth, plus a homepage and topics
Licensing: 0%; How freely the code can be adopted

What to know before adopting

No release in over a year
No license — all rights reserved by default, not safe to reuse
No licence file — default copyright applies, all rights reserved.

Details

Created: 10 years ago
Last push: 17 days ago
Last release: 9 years ago
Release tag: 1.1.1000
Releases in the last year: 0
Default branch: master
Size: 685k KB
Owner type: User
Owner location: —
Discovered via: phrase:computer vision
First indexed: 7 days ago
Last synced: 1 hour ago

jiafeng5513/Evision

计算机视觉实践和探索/Practice and explorations in computer vision.

52No licenceC++

View on GitHub

Stars: 438
Forks: 154
Watchers: 16
Open issues: 5
Contributors: —

Star history, last 90 days— over 2 snapshots

2026-07-31 · 4382026-08-06 · 438

Star history, last 90 days
Date	Stars
2026-07-31	438
2026-08-06	438

Today

— stars today

This week

— stars this week

This month

— stars this month

Momentum

0.0

growth rate 0.00%/day

From the README

<div align=center><img width="100" height="100" src="./src/EvisionSandbox/resource/Evision.ico"/></div>

<div align=center>Evision 双目视觉系统</div>
<div align=center><a href="https://www.bilibili.com/video/av46024738">演示视频</a></div>
<div align=center></div>
<div align=center>如果您觉得有帮助,请为该项目点star.以便于及时收到最新更新.</div>

## 重要提示:Evision 已迁移至 Qt6 + vcpkg 工具链

> **复活与现代化进行中**:本项目正在借助 AI 工具进行阶段性复活与现代化重构。阶段 0(仓库清理)与阶段 1(让 master 重新可编译、CMake 现代化、内存泄漏修复)已完成;内嵌 Darknet 目标检测模块已整体移除(归档于 `archive/darknet-fork` 分支)。后续阶段将引入深度学习立体匹配等 AI 能力。详见下方"路线图"。

Introduction:
=========
1. 双目系统的标定,畸变校正,视差,三维重建,距离测量等.<br>
2. ELAS,ADCensus视差算法.<br>
3. 关于双目的中文资料重复度太高,希望各位后来者能够吸取前人精华,摒弃前人的糟粕,多多自行探索,不要抄来抄去.<br>
4. 关于程序的BUG,以及其他困惑,请使用issues,或联系邮件[email protected].<br>
5. Evision使用 cmake-vcpkg 工具链进行编译,开发者使用 Visual Studio 2019+,Qt6,CMake 3.20+.<br>

资助作者
========
<div align=center><img width="500" height="339" src="./doc/pay.png"/></div>


目录
========
- [Introduction:](#introduction)
- [资助作者](#资助作者)
- [目录](#目录)
      - [1.Dependencies](#1dependencies)
      - [2.Directory_specification](#2directory_specification)
      - [3.build](#3build)
      - [4.模块说明](#4模块说明)
      - [5.双目设备](#5双目设备)
      - [6.参考文献](#6参考文献)
      - [7.数据下载](#7数据下载)
      - [8.路线图](#8路线图)

#### 1.路径说明
1. `data`文件夹存储测试用例(大体积数据需从 Release 下载,见下文).<br>
2. `doc`文件夹中存储的是文档和文档中所用的图片等资源,提供了本项目使用的一些算法相关的论文.<br>
3. `scripts`文件夹存储编译辅助脚本.<br>
4. `src`文件夹中存放代码文件.<br>
* ~~`legacy`~~ 和 ~~`install`~~ 目录已在新一轮清理中移除(旧 C# 版本与编译产物归档于 `archive/legacy-pre-cleanup` 分支).<br>
* ~~`package`~~ 目录已移除(pthread windows 版改由 vcpkg 管理).<br>

#### 2.编译基础组件
1. 安装 Visual Studio 2019 或更高版本 + 使用C++的桌面开发工作负载 + 英文语言包.([visual studio Installer](https://visualstudio.microsoft.com/zh-hans/vs/))<br>
2. 安装 CMake 3.20+.([Download CMake](https://cmake.org/download/))<br>
3. 安装 vcpkg.([GitHub For vcpkg](https://github.com/Microsoft/vcpkg))<br>
4. `vcpkg install opencv:x64-windows`<br>
5. `vcpkg install boost:x64-windows`<br>
6. `vcpkg install pcl:x64-windows`<br>
7. `vcpkg install qt6:x64-windows`<br>
8. `mkdir sln,cd sln`<br>
9. `cmake ..\src\ -DCMAKE_TOOLCHAIN_FILE="${vcpkg_root}/scripts/buildsystems/vcpkg.cmake" -G"Visual Studio 16 2019" -A x64 -DPointCloudViewer=ON`<br>
10. 使用 Visual Studio 打开 `Evision.sln`.<br>
11. 右键 `_CMakeTargets/ALL_BUILD`,生成.<br>
12. 右键 `EvisionSandBox`,设为启动项目.<br>
13. 点击"本地Windows调试器",启动程序.<br>
* 如果vcpkg下载包的速度过慢,可以上代理(以PowerShell为例)`$env:HTTP_PROXY="http://127.0.0.1:1080"`

#### 3.模块说明   
|      |模块名|功能|UI|输出目标|状态|
|:----:|:----:|:----:|:----:|:----:|:----:|
| 1|EvisionADCensus           |ADCensus视差算法  | ×    |动态链接库|✅|
| 2|EvisionElas               |Elas视差算法      | ×    |动态链接库|✅|
| 3|EvisionPnP                |PnP              | ×    |动态链接库|✅|
| 4|EvisionMonocularCalib     |单目标定          | √    |动态链接库|✅|
| 5|EvisionPolyTracker        |单目几何体追踪     | √    |动态链接库|✅|
| 6|EvisionCalibrate          |双目标定          | √    |动态链接库|✅|
| 7|EvisionDisparity          |视差(立体匹配)    |√   |动态链接库|✅|
| 8|EvisionTrace              |交互式测量        | √   |动态链接库|✅|
| 9|EvisionUndistortion       |畸变校正          | √   |动态链接库|✅|  
|10|EvisionCamera             |单目和双目相机功能 | √    |动态链接库|✅|   
|11|EvisionCloudViewer        |三维点云查看      | √    |动态链接库|✅|
|12|EvisionParamBridge        |外部参数传递      | √   |动态链接库|✅|  
|13|EvisionUtils              |通用工具类        | ×   |动态链接库|✅|    
|14|EvisionSandbox            |主程序UI          | √    |可执行程序|✅|

> **EvisionCamera 模块**:已迁移至 Qt6 Multimedia API(`QCameraDevice`/`QImageCapture`/`QMediaCaptureSession`),使用 `QVideoWidget` 替代已移除的 `QCameraViewfinder`。编解码器选择 UI 保留为占位(Qt6 `QImageCapture` 不再暴露编解码器 API)。

#### 4.双目设备
1. 需要注意的是,视差效果,点云效果和精度和设备关系非常大,图片的分辨率越高,光照条件越好,畸变越小,一致性越好,最终效果也就越好.此外,两个相机的距离(基线长度)会影响系统的有效范围,一般来讲,基线越长的双目系统越容易获取远处目标的视差,延长基线能够一定程度上(因为有效距离和相机的焦段也有关)将系统的有效范围拉得更远,但同时,基线越长,盲区(距离相机过近的目标不会同时出现在两个视野中)也越大.
2. 推荐的双目系统:
   1. ZED/RealSense/MYNTEYE小觅相机.这是成熟(昂贵)的商业产品,出厂带有高精度的标定数据和功能强大的SDK,而且还带有IMU,IR主动光学等辅助设备,适合做SLAM,笔者认为购买这类相机是最节约时间成本的方法.<br>
   ![image](./doc/cameras.png)
   1. 双目开发板.淘宝上有很多这类产品,价格比ZED那类便宜很多,同时他们

Excerpt of 7,920 characters

Read on GitHub

Top contributors

No contributor data yet — it is fetched during the contributor sync pass.

Quality score

Would you bet a product on this? Bounded 0–100 and slow moving.

52/10052

Maintenance: 90%; How recently the default branch was pushed to
Releases: 10%; Recency and cadence of tagged releases
Community: 20%; Contributor breadth and bus factor
Issue backlog: 85%; Open issues measured against the size of the audience
Documentation: 83%; README depth, plus a homepage and topics
Licensing: 0%; How freely the code can be adopted

What to know before adopting

No release in over a year

Details

Created: 10 years ago
Last push: 17 days ago
Last release: 9 years ago
Release tag: 1.1.1000
Releases in the last year: 0
Default branch: master
Size: 685k KB
Owner type: User
Owner location: —
Discovered via: phrase:computer vision
First indexed: 7 days ago
Last synced: 1 hour ago

Top AI Repos — open-source AI, indexed and scored

jiafeng5513/Evision

From the README

Top contributors

Quality score

What to know before adopting

Categories

Details

jiafeng5513/Evision

From the README

Top contributors

Quality score

What to know before adopting

Categories

Details

Top AI Repos — open-source AI, indexed and scored

Trend

From the README

Top contributors

Quality score

What to know before adopting

Categories

Details

Trend

From the README

Top contributors

Quality score

What to know before adopting

Categories

Details