https://github.com/sunyancn/image-tucker-compress

Image Tucker Compress
https://github.com/sunyancn/image-tucker-compress

image tucker-decomposition

Last synced: 3 months ago
JSON representation

Image Tucker Compress

• Host: GitHub
• URL: https://github.com/sunyancn/image-tucker-compress
• Owner: SunYanCN
• License: mit
• Created: 2018-09-11T07:07:59.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2018-09-11T07:49:51.000Z (over 6 years ago)
• Last Synced: 2025-01-12T06:49:10.772Z (4 months ago)
• Topics: image, tucker-decomposition
• Language: Python
• Size: 1.05 MB
• Stars: 2
• Watchers: 1
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Image-Tucker-Compress

分别使用tensorly和tensorcomlib测试了张量图像压缩

## 安装Python库

`numpy,pillow,matplotlib,tensorly,skimage,tensorcomlib`

## Tensorly测试结果

运行`tensorly`文件夹下的`image_tucker.py`文件，设置`tucker_rank`分别为[50,50,3]、[100,100,3]、[200,200,3]和[300,300,3]。如果报`Memory`错误，修改`tensorly`源代码，`full_matrx = False`。

运行结果如下：

### rank = [50,50,3]

![](https://i.loli.net/2018/09/11/5b976d893cab0.png)

```

Image Compression Ratio：0.925347646077474

Image Compare PSNR：25.946412228266816

Decomposition Time：0.7083556652069092

```

### rank = [100,100,3]

![](https://i.loli.net/2018/09/11/5b976da997704.png)

```

Image Compression Ratio：0.8316332499186198

Image Compare PSNR：28.686432064795255

Decomposition Time：0.8138234615325928

```

### rank = [200,200,3]

![](https://i.loli.net/2018/09/11/5b976dc42f20a.png)

```

Image Compression Ratio：0.5869839986165364

Image Compare PSNR：30.804968069749638

Decomposition Time：1.4251620769500732

```

### rank = [300,300,3]

![](https://i.loli.net/2018/09/11/5b976de410a5f.png)

```

Image Compression Ratio：0.2660408020019531

Image Compare PSNR：38.38228135960186

Decomposition Time：1.942857027053833

```

## Tensorcomlib

Tensorcomlib是我自己练习写的一个库，测试效果和tensorly相似，输出信息可能更详细一点。

### 比较

![](https://i.loli.net/2018/09/11/5b976e62db9fd.png)

## 联系方式

[email protected]

## 参考资料

1. Tensorly文档

2. 《基于张量Tucker分解的彩色图像压缩 》，王东方