# Update v0.1.12
- Support text region with different angle.
- Higher resolution. min(height, width) of output is 256 (128 in v0.24.0)
See our project page for more details: https://github.com/csxmli2016/textbsr
-----------
## This is a simple text image super-resolution package.
More details can be found in our Project Page: https://github.com/csxmli2016/textbsr
This package can post-process the text region with a simple command, i.e.,
```
textbsr -i [LR_TEXT_PATH] -b [BACKGROUND_SR_PATH]
```
> - [LR_TEXT_PATH] is the LR image path.
> - [BACKGROUND_SR_PATH] stores the results from any blind image super-resolution methods.
> - If the text image is degraded severely, this method may still fail to obtain a plausible result.
### Dependencies and Installation
- numpy
- cnstd
- torch>=1.8.1
- torchvision>=0.9
```
# Install with pip
pip install textbsr
```
### Basic Usage
```
# On the terminal command
textbsr -i [LR_TEXT_PATH]
```
or
```
# On the python environment
from textbsr import textbsr
textbsr.bsr(input_path='./testsets/LQs')
```
Parameter details:
| parameter name | default | description |
| :----- | :-----: | :----- |
| <span style="white-space:nowrap">-i, --input_path </span>| - | The lr text image path. It can store full images or text layouts only. |
| <span style="white-space:nowrap">-b, --bg_path</span> | None | The background sr path from other methods. If None, we only restore the text region detected by cnstd.|
| <span style="white-space:nowrap">-o, --output_path</span> | None | The save path for text sr result. If None, we save the results on the same path with the format of [input_path]\_TIMESTAMP.|
| <span style="white-space:nowrap">-a, --aligned </span>| False | action='store_true'. If True, the input text image contains only text region. If False, we use CnSTD to detect text regions and then restore them.|
| <span style="white-space:nowrap">-s, --save_text </span>| False | action='store_true'. If True, save the LR and SR text layout.|
| <span style="white-space:nowrap">-d, --device</span> | None | Device, use 'gpu' or 'cpu'. If None, we use torch.cuda.is_available to select the device. |
### Example for post-processing the text region
```
# On the terminal command
textbsr -i [LR_TEXT_PATH] -b [BACKGROUND_SR_PATH] -s
```
or
```
# On the python environment
from textbsr import textbsr
textbsr.bsr(input_path='./testsets/LQs', bg_path='./testsets/RealESRGANResults', save_text=True)
```
> When [BACKGROUND_SR_PATH] is None, we only restore the text region and paste it back to the LR input, with the background region unchanged.
---
### Example for restoring the aligned text region
```
# On the terminal command
textbsr -i [LR_TEXT_PATH] -a
```
or
```
# On the python environment
from textbsr import textbsr
textbsr.bsr(input_path='./testsets/LQs', aligned=True)
```
> If you find this package helpful, please kindly consider citing our paper:
```
@InProceedings{li2023marconet,
author = {Li, Xiaoming and Zuo, Wangmeng and Loy, Chen Change},
title = {Learning Generative Structure Prior for Blind Text Image Super-resolution},
booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
year = {2023}
}
```
Raw data
{
"_id": null,
"home_page": "https://github.com/csxmli2016/MARCONet",
"name": "textbsr",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.6",
"maintainer_email": "",
"keywords": "blind text image super-resolution",
"author": "Xiaoming Li",
"author_email": "csxmli@gmail.com",
"download_url": "https://files.pythonhosted.org/packages/83/c3/0810ba664432e956270810e935ed5bf7252c3a417bd333dbfbdac7eed3d1/textbsr-0.1.12.tar.gz",
"platform": null,
"description": "\n# Update v0.1.12\n\n- Support text region with different angle.\n- Higher resolution. min(height, width) of output is 256 (128 in v0.24.0)\n\nSee our project page for more details: https://github.com/csxmli2016/textbsr\n\n-----------\n\n## This is a simple text image super-resolution package.\nMore details can be found in our Project Page: https://github.com/csxmli2016/textbsr\n\nThis package can post-process the text region with a simple command, i.e., \n```\ntextbsr -i [LR_TEXT_PATH] -b [BACKGROUND_SR_PATH]\n```\n\n> - [LR_TEXT_PATH] is the LR image path.\n> - [BACKGROUND_SR_PATH] stores the results from any blind image super-resolution methods.\n> - If the text image is degraded severely, this method may still fail to obtain a plausible result.\n\n### Dependencies and Installation\n- numpy\n- cnstd\n- torch>=1.8.1\n- torchvision>=0.9\n\n``` \n# Install with pip\npip install textbsr\n```\n\n\n### Basic Usage\n\n```\n# On the terminal command\ntextbsr -i [LR_TEXT_PATH]\n```\nor\n```\n# On the python environment\nfrom textbsr import textbsr\ntextbsr.bsr(input_path='./testsets/LQs')\n```\n\nParameter details:\n\n| parameter name | default | description |\n| :----- | :-----: | :----- |\n| <span style=\"white-space:nowrap\">-i, --input_path </span>| - | The lr text image path. It can store full images or text layouts only. |\n| <span style=\"white-space:nowrap\">-b, --bg_path</span> | None | The background sr path from other methods. If None, we only restore the text region detected by cnstd.|\n| <span style=\"white-space:nowrap\">-o, --output_path</span> | None | The save path for text sr result. If None, we save the results on the same path with the format of [input_path]\\_TIMESTAMP.|\n| <span style=\"white-space:nowrap\">-a, --aligned </span>| False | action='store_true'. If True, the input text image contains only text region. If False, we use CnSTD to detect text regions and then restore them.|\n| <span style=\"white-space:nowrap\">-s, --save_text </span>| False | action='store_true'. If True, save the LR and SR text layout.|\n| <span style=\"white-space:nowrap\">-d, --device</span> | None | Device, use 'gpu' or 'cpu'. If None, we use torch.cuda.is_available to select the device. |\n\n### Example for post-processing the text region\n```\n# On the terminal command\ntextbsr -i [LR_TEXT_PATH] -b [BACKGROUND_SR_PATH] -s\n```\nor\n```\n# On the python environment\nfrom textbsr import textbsr\ntextbsr.bsr(input_path='./testsets/LQs', bg_path='./testsets/RealESRGANResults', save_text=True)\n```\n> When [BACKGROUND_SR_PATH] is None, we only restore the text region and paste it back to the LR input, with the background region unchanged.\n\n\n---\n\n### Example for restoring the aligned text region\n```\n# On the terminal command\ntextbsr -i [LR_TEXT_PATH] -a\n```\nor\n```\n# On the python environment\nfrom textbsr import textbsr\ntextbsr.bsr(input_path='./testsets/LQs', aligned=True)\n```\n\n\n\n> If you find this package helpful, please kindly consider citing our paper:\n```\n@InProceedings{li2023marconet,\nauthor = {Li, Xiaoming and Zuo, Wangmeng and Loy, Chen Change},\ntitle = {Learning Generative Structure Prior for Blind Text Image Super-resolution},\nbooktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},\nyear = {2023}\n}\n```\n\n\n",
"bugtrack_url": null,
"license": "S-Lab License 1.0",
"summary": "a simple version for blind text image super-resolution (current version is only for English and Chinese)",
"version": "0.1.12",
"project_urls": {
"Homepage": "https://github.com/csxmli2016/MARCONet"
},
"split_keywords": [
"blind",
"text",
"image",
"super-resolution"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "da0b267d3298fe0f84cc7ecf24746678ba69622fca629d160765fb21389404e2",
"md5": "13d5636038ebdd66fda05ad4c4bc9648",
"sha256": "adaa1799259b5a18b84efaded1523188ddc763928003cb639507ef756d5c19dd"
},
"downloads": -1,
"filename": "textbsr-0.1.12-py3-none-any.whl",
"has_sig": false,
"md5_digest": "13d5636038ebdd66fda05ad4c4bc9648",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.6",
"size": 10875,
"upload_time": "2023-06-05T13:29:18",
"upload_time_iso_8601": "2023-06-05T13:29:18.235727Z",
"url": "https://files.pythonhosted.org/packages/da/0b/267d3298fe0f84cc7ecf24746678ba69622fca629d160765fb21389404e2/textbsr-0.1.12-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "83c30810ba664432e956270810e935ed5bf7252c3a417bd333dbfbdac7eed3d1",
"md5": "0147b332cf040ef3ec7395c43cb6da6a",
"sha256": "5dc71fec4119d5e6f656679f7adaf55b9d390ec8b4de508604461d3c050f92d8"
},
"downloads": -1,
"filename": "textbsr-0.1.12.tar.gz",
"has_sig": false,
"md5_digest": "0147b332cf040ef3ec7395c43cb6da6a",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.6",
"size": 10256,
"upload_time": "2023-06-05T13:29:19",
"upload_time_iso_8601": "2023-06-05T13:29:19.787464Z",
"url": "https://files.pythonhosted.org/packages/83/c3/0810ba664432e956270810e935ed5bf7252c3a417bd333dbfbdac7eed3d1/textbsr-0.1.12.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-06-05 13:29:19",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "csxmli2016",
"github_project": "MARCONet",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [],
"lcname": "textbsr"
}
JSON
