Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/YanjieZe/NU-SSSR
NU-SSSR (Non-Uniformly Sparsely Sampled Super Resolution). 非均匀稀疏样本图像超分辨率重建:基于傅里叶变换撒点法与Voronoi-Delaunay三角化的高层视觉与低层视觉并用方法
https://github.com/YanjieZe/NU-SSSR
Last synced: 6 days ago
JSON representation
NU-SSSR (Non-Uniformly Sparsely Sampled Super Resolution). 非均匀稀疏样本图像超分辨率重建:基于傅里叶变换撒点法与Voronoi-Delaunay三角化的高层视觉与低层视觉并用方法
- Host: GitHub
- URL: https://github.com/YanjieZe/NU-SSSR
- Owner: YanjieZe
- Created: 2021-10-29T13:27:52.000Z (about 3 years ago)
- Default Branch: master
- Last Pushed: 2022-01-02T07:17:35.000Z (almost 3 years ago)
- Last Synced: 2024-07-27T15:14:34.726Z (4 months ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 29.9 MB
- Stars: 9
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- awesome-cs - @YanjieZe @Purewhite2019, 2021 Fall
README
# 非均匀稀疏样本图像超分辨率重建:基于傅里叶变换撒点法与Voronoi-Delaunay三角化的高层视觉与低层视觉并用方法
```
Non-uniformly Sparsely Sampled Image Super-Resolution and Reconstruction: A Method Based on Fourier Transformation Point Sampling and Voronoi-Delaunay Triangulation from High Level Vision and Low Level Vision
```
迮炎杰,刘祺# 项目说明
本项目为**CS337:计算机图形学**的课程大作业。课题为“非均匀稀疏样本图像超采样”。我们的算法框架主要致力于解决如下问题:将一张非均匀采样的图像尽可能复原为原图。
下面展示了使用我们的算法进行非均匀采样后的图像。这三张图分别为:原图,使用1000个采样点的图像,使用10000个采样点的图像。
![](imgs/img_origin.jpeg)![](imgs/blur_point1000.jpg)
![](imgs/blur_point10000.png)
# 系统框架
我们的算法框架支持多种采样方法,多种着色方法,多种SOTA的神经网络算法。下图展示了我们的算法的整体流程。
![](imgs/overall.png)# 使用方法
1. 训练各类CG模型:`sh scripts/train_cg.sh`
2. 测试各类CG模型:`sh scripts/eval_cg.sh`想要使用不同的模型(SRCNN,SRCNN2,SwinIR,VDSR,MAE)与不同参数请在对应script中修改。