GitPlanet

Cheap and reliable Node.js hosting starts at $3/month, and $1/month static HTML hosting

Created with love in Canada, visit hostnodejs.com today

Feel like to post an Ad? Learn Details
All Projects → ethan-li-coding → Patchmatchstereo

ethan-li-coding / Patchmatchstereo

Licence: mit
PatchMatch，倾斜窗口经典，效果极佳，OpenMVS&Colmap稠密匹配算法。完整实现，代码规范，注释清晰，博客教学，欢迎star！

Labels

3d-reconstruction

Projects that are alternatives of or similar to Patchmatchstereo

Prnet pytorch
Training & Inference Code of PRNet in PyTorch 1.1.0
Stars: ✭ 149 (-31.96%)
Mutual labels:  3d-reconstruction
Meshlab
The open source mesh processing system
Stars: ✭ 2,619 (+1095.89%)
Mutual labels:  3d-reconstruction
Horizonnet
Pytorch implementation of HorizonNet: Learning Room Layout with 1D Representation and Pano Stretch Data Augmentation.
Stars: ✭ 195 (-10.96%)
Mutual labels:  3d-reconstruction
Flame pytorch
This is a implementation of the 3D FLAME model in PyTorch
Stars: ✭ 153 (-30.14%)
Mutual labels:  3d-reconstruction
Factored3d
Code release for "Factoring Shape, Pose, and Layout from the 2D Image of a 3D Scene"
Stars: ✭ 170 (-22.37%)
Mutual labels:  3d-reconstruction
Gan2shape
Code for GAN2Shape (ICLR2021 oral)
Stars: ✭ 183 (-16.44%)
Mutual labels:  3d-reconstruction
Drc
Code release for "Multi-view Supervision for Single-view Reconstruction via Differentiable Ray Consistency" (CVPR 2017)
Stars: ✭ 147 (-32.88%)
Mutual labels:  3d-reconstruction
Scancomplete
[CVPR'18] ScanComplete: Large-Scale Scene Completion and Semantic Segmentation for 3D Scans
Stars: ✭ 213 (-2.74%)
Mutual labels:  3d-reconstruction
Opensfm
Open source Structure-from-Motion pipeline
Stars: ✭ 2,342 (+969.41%)
Mutual labels:  3d-reconstruction
Tf flame
Tensorflow framework for the FLAME 3D head model. The code demonstrates how to sample 3D heads from the model, fit the model to 2D or 3D keypoints, and how to generate textured head meshes from Images.
Stars: ✭ 193 (-11.87%)
Mutual labels:  3d-reconstruction
Flame
FLaME: Fast Lightweight Mesh Estimation
Stars: ✭ 164 (-25.11%)
Mutual labels:  3d-reconstruction
Deodr
A differentiable 3D renderer with Pytorch, Tensorflow and Matlab interfaces
Stars: ✭ 167 (-23.74%)
Mutual labels:  3d-reconstruction
Pixel2meshplusplus
Pixel2Mesh++: Multi-View 3D Mesh Generation via Deformation. In ICCV2019.
Stars: ✭ 188 (-14.16%)
Mutual labels:  3d-reconstruction
Sltk
An OpenCV-based structured light processing toolkit.
Stars: ✭ 151 (-31.05%)
Mutual labels:  3d-reconstruction
Computervisiondatasets
Stars: ✭ 207 (-5.48%)
Mutual labels:  3d-reconstruction
Surfelwarp
SurfelWarp: Efficient Non-Volumetric Dynamic Reconstruction
Stars: ✭ 149 (-31.96%)
Mutual labels:  3d-reconstruction
Tailornet
Code for our CVPR 2020 (ORAL) paper - TailorNet: Predicting Clothing in 3D as a Function of Human Pose, Shape and Garment Style.
Stars: ✭ 180 (-17.81%)
Mutual labels:  3d-reconstruction
Pix2vox
Implementation of "Pix2Vox: Context-aware 3D Reconstruction from Single and Multi-view Images" (Xie et al., ICCV 2019)
Stars: ✭ 216 (-1.37%)
Mutual labels:  3d-reconstruction
3dreconstruction
3D reconstruction, sfm with Python3
Stars: ✭ 213 (-2.74%)
Mutual labels:  3d-reconstruction
Msn Point Cloud Completion
Morphing and Sampling Network for Dense Point Cloud Completion (AAAI2020)
Stars: ✭ 196 (-10.5%)
Mutual labels:  3d-reconstruction
View All Similar Projects ➔

PatchMatchStereo

PatchMatch，倾斜窗口经典，效果极佳，OpenMVS&Colmap稠密匹配算法。完整实现，代码规范，注释清晰，博客教学，欢迎star！

CSDN博客

如果你想了解更多代码讲解，下面的博客或许能帮到你：
【码上实战】【立体匹配系列】经典PatchMatch: （1）框架
【码上实战】【立体匹配系列】经典PatchMatch: （2）主类
【码上实战】【立体匹配系列】经典PatchMatch: （3）随机初始化
【码上实战】【立体匹配系列】经典PatchMatch: （4）代价计算
【码上实战】【立体匹配系列】经典PatchMatch: （5）迭代传播
【码上实战】【立体匹配系列】经典PatchMatch: （6）后处理

环境

windows10 / visual studio 2015&2019
代码基本没有使用系统api，你可以非常方便的移植到linux，可能需要做极少量的修改

强烈建议你使用release模式运行代码，强烈不建议使用debug模式运行代码

第三方库

opencv310
百度网盘连接：https://pan.baidu.com/s/1_WD-KdPyDBazEIim7NU3jA
提取码：aab4

解压后放将名称为OpenCV的文件夹复制到到3rdparty文件夹下

若运行时提示缺少opencv_world310(d).dll，则在OpenCV文件夹里找到对应的dll文件复制到程序exe所在的目录即可（Opencv\dll\opencv_world310(d).dll），带d为debug库，不带d为release库。

为便于移植，算法是不依赖任何图像库的，只在算法实验部分调用opencv库读取和显示图像，也可替换成其他图像库

算法引导

算法框架图



备注

算法优点：效果好，内存占用低
算法缺点：效率低，速度比较慢，不建议跑大图，建议跑个小图看看效果（Release模式）。如果设置为前端平行窗口（PatchMatchStereo为倾斜窗口时效果最好），则速度会更快，如下：

pms_option.is_fource_fpw = true;

论文

Bleyer M, Rhemann C, Rother C. PatchMatch Stereo-Stereo Matching with Slanted Support Windows[C]. British Machine Vision Conference 2011. 2011.

Github图片不显示的解决办法

修改hosts

C:\Windows\System32\drivers\etc\hosts

在文件末尾添加：

# GitHub Start
192.30.253.119      gist.github.com
151.101.184.133     assets-cdn.github.com
151.101.184.133     raw.githubusercontent.com
151.101.184.133     gist.githubusercontent.com
151.101.184.133     cloud.githubusercontent.com
151.101.184.133     camo.githubusercontent.com
151.101.184.133     avatars0.githubusercontent.com
151.101.184.133     avatars1.githubusercontent.com
151.101.184.133     avatars2.githubusercontent.com
151.101.184.133     avatars3.githubusercontent.com
151.101.184.133     avatars4.githubusercontent.com
151.101.184.133     avatars5.githubusercontent.com
151.101.184.133     avatars6.githubusercontent.com
151.101.184.133     avatars7.githubusercontent.com
151.101.184.133     avatars8.githubusercontent.com
# GitHub End
代码框架图
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].