mirror of
https://github.com/LCTT/TranslateProject.git
synced 2024-12-26 21:30:55 +08:00
commit
7fb90424a5
@ -1,77 +0,0 @@
|
||||
[#]: subject: "Edit your photos with open source artificial intelligence"
|
||||
[#]: via: "https://opensource.com/article/23/4/edit-photos-open-source-ai"
|
||||
[#]: author: "Don Watkins https://opensource.com/users/don-watkins"
|
||||
[#]: collector: "lkxed"
|
||||
[#]: translator: "geekpi"
|
||||
[#]: reviewer: " "
|
||||
[#]: publisher: " "
|
||||
[#]: url: " "
|
||||
|
||||
Edit your photos with open source artificial intelligence
|
||||
======
|
||||
|
||||
I've been interested in photography ever since I co-opted my father's Kodak 620 camera as a young boy. I used it to take pictures of the flora and fauna of our neighborhood. My love of photography led me to an Instamatic camera in high school, and eventually to digital cameras as they entered the marketplace in the late 1990s. Early digital cameras provided portability and the ability to quickly capture and easily share images on the internet. But they lacked the quality and complexity of the best of film photography. Of course digital cameras have improved a lot since then. But I have years of digital photographs that just look a little, well, _little_ on modern devices.
|
||||
|
||||
Until recently, my go-to tool for upscaling digital images has been [GIMP][1]. A couple of years ago, I tried to use GIMP to upscale a thumbnail image of my father that was taken in the mid-1940s. It worked, but the photo lacked the detail, depth, and clarity that I wanted.
|
||||
|
||||
That's all changed since I learned about [Upscayl][2], a free and open source program that uses [open source artificial intelligence][3] to upscale low-resolution images.
|
||||
|
||||
### Upscayl
|
||||
|
||||
Upscayl works on Linux, Windows, and macOS.
|
||||
|
||||
It's easy to install on Linux whether your system uses RPM or DEB packages, and its website contains a universal Linux [AppImage][4] too.
|
||||
|
||||
For macOS and Windows, you can download installers from the project's website. Upscayl is released with an [AGPL][5] license.
|
||||
|
||||
### Get started with Upscayl
|
||||
|
||||
Once installed, you can begin upscaling your images. The GUI software is very easy to use. The software makes your old images look like they were taken yesterday with image resolutions that far exceed the originals. In addition, you can batch scale entire folders and photo albums of images and upscale them all at once.
|
||||
|
||||
![The left panel of Upscayl provides clear guidance on the 4 steps required to upscale an image.][6]
|
||||
|
||||
Launch the software and click the **Select Image** button. Find the image or folder of images you want to upscale.
|
||||
|
||||
Once the image is loaded, select the type of upscaling you want to try. The default is Real-ESRGAN, and that's a good place to start. There are six options to choose from, including a selection for digital art.
|
||||
|
||||
- General photo with [Real-ESRGAN][7]
|
||||
- General photo with [remacri][8]
|
||||
- General photo with [ultramix balanced][8]
|
||||
- General photo with [ultrasharp][8]
|
||||
- Digital Art
|
||||
- Sharpen Image
|
||||
|
||||
Next, select the output directory where you want your upscaled images to be saved.
|
||||
|
||||
And finally, click the **Upscayl** button to begin the upscaling process. The speed of conversion depends on your GPU and the image output choice you make.
|
||||
|
||||
Here's a test image, with the low-resolution image on the left and the Upscayl version on the right:
|
||||
|
||||
![An image processed by Upscayl.][9]
|
||||
|
||||
### Time to try Upscayl for your images
|
||||
|
||||
Upscayl is one of my favorite upscaling applications. It does depend heavily on your GPU, so it may not work on an old computer or one with an especially weak graphics card. But there's no harm in giving it a try. So download it and try it. I think you'll be impressed with the results.
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://opensource.com/article/23/4/edit-photos-open-source-ai
|
||||
|
||||
作者:[Don Watkins][a]
|
||||
选题:[lkxed][b]
|
||||
译者:[译者ID](https://github.com/译者ID)
|
||||
校对:[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译,[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]: https://opensource.com/users/don-watkins
|
||||
[b]: https://github.com/lkxed/
|
||||
[1]: https://opensource.com/tags/gimp
|
||||
[2]: https://github.com/upscayl/upscayl
|
||||
[3]: https://opensource.com/article/22/10/defining-open-source-ai
|
||||
[4]: https://appimage.github.io/Upscayl/
|
||||
[5]: https://github.com/upscayl/upscayl/blob/main/LICENSE
|
||||
[6]: https://opensource.com/sites/default/files/2023-03/upscayl-panel.webp
|
||||
[7]: https://github.com/xinntao/Real-ESRGAN
|
||||
[8]: https://upscale.wiki/wiki/Model_Database
|
||||
[9]: https://opensource.com/sites/default/files/2023-03/jurica-koletic-7YVZYZeITc8-unsplash.webp
|
@ -0,0 +1,77 @@
|
||||
[#]: subject: "Edit your photos with open source artificial intelligence"
|
||||
[#]: via: "https://opensource.com/article/23/4/edit-photos-open-source-ai"
|
||||
[#]: author: "Don Watkins https://opensource.com/users/don-watkins"
|
||||
[#]: collector: "lkxed"
|
||||
[#]: translator: "geekpi"
|
||||
[#]: reviewer: " "
|
||||
[#]: publisher: " "
|
||||
[#]: url: " "
|
||||
|
||||
使用开源人工智能编辑你的照片
|
||||
======
|
||||
|
||||
自从我小时候使用了我父亲的柯达 620 相机以来,我就一直对摄影很感兴趣。我用它来拍摄我们附近的动植物。我对摄影的热爱使我在高中时购买了一台 Instamatic 相机,并最终在 1990 年代后期数码相机进入市场时购买了一台。早期的数码相机具有便携性,能够快速捕捉图像并轻松地在互联网上共享图像。但它们缺乏最好的胶片摄影的质量和复杂性。当然,从那时起,数码相机有了很大的改进。但我有多年的数码照片,在现代设备上看起来有点_小_。
|
||||
|
||||
直到最近,我用于升级数字图像的首选工具一直是 [GIMP][1]。几年前,我尝试使用 GIMP 放大我父亲在 1940 年代中期拍摄的缩略图。它奏效了,但照片缺乏我想要的细节、深度和清晰度。
|
||||
|
||||
自从我了解了 [Upscayl][2],这是一个免费的开源程序,它使用[开源人工智能][3]来升级低分辨率图像,这一切都发生了变化。
|
||||
|
||||
### Upscayl
|
||||
|
||||
Upscayl 适用于 Linux、Windows 和 macOS。
|
||||
|
||||
无论你的系统使用 RPM 还是 DEB 包,在 Linux 上安装都很容易,它的网站也包含一个通用的 Linux [AppImage][4]。
|
||||
|
||||
对于 macOS 和 Windows,你可以从项目的网站下载安装程序。Upscayl 使用 [AGPL][5] 许可证发布。
|
||||
|
||||
### 开始使用 Upscayl
|
||||
|
||||
安装后,你可以放大图像。GUI 软件非常易于使用。该软件使你的旧图像看起来像昨天拍摄的,图像分辨率远远超过原件。此外,你可以批量缩放整个文件夹和图像相册,并一次对它们进行升级。
|
||||
|
||||
![The left panel of Upscayl provides clear guidance on the 4 steps required to upscale an image.][6]
|
||||
|
||||
启动软件并单击**选择图像**按钮。找到要放大的图像或图像文件夹。
|
||||
|
||||
加载图像后,选择要尝试的放大类型。默认值为 Real-ESRGAN,这是一个很好的起点。有六个选项可供选择,包括数字艺术的选择。
|
||||
|
||||
- 使用 [Real-ESRGAN][7] 的一般照片
|
||||
- 使用 [remacri][8] 的一般照片
|
||||
- 使用 [ultramix balanced][8] 的一般照片
|
||||
- 使用 [ultrasharp][8] 的一般照片
|
||||
- 数字艺术
|
||||
- 锐化图像
|
||||
|
||||
接下来,选择要保存放大图像的输出目录。
|
||||
|
||||
最后,单击 **Upscayl** 按钮开始放大过程。转换速度取决于你的 GPU 和你选择的图像输出。
|
||||
|
||||
这是一张测试图像,左侧是低分辨率图像,右侧是 Upscayl 版本:
|
||||
|
||||
![An image processed by Upscayl.][9]
|
||||
|
||||
### 是时候为你的图像尝试 Upscayl 了
|
||||
|
||||
Upscayl 是我最喜欢的放大应用之一。它确实在很大程度上取决于你的 GPU,因此它可能无法在旧计算机或显卡特别弱的计算机上运行。但是尝试一下也没有坏处。所以下载并尝试一下。我想你会对结果印象深刻。
|
||||
|
||||
--------------------------------------------------------------------------------
|
||||
|
||||
via: https://opensource.com/article/23/4/edit-photos-open-source-ai
|
||||
|
||||
作者:[Don Watkins][a]
|
||||
选题:[lkxed][b]
|
||||
译者:[geekpi](https://github.com/geekpi)
|
||||
校对:[校对者ID](https://github.com/校对者ID)
|
||||
|
||||
本文由 [LCTT](https://github.com/LCTT/TranslateProject) 原创编译,[Linux中国](https://linux.cn/) 荣誉推出
|
||||
|
||||
[a]: https://opensource.com/users/don-watkins
|
||||
[b]: https://github.com/lkxed/
|
||||
[1]: https://opensource.com/tags/gimp
|
||||
[2]: https://github.com/upscayl/upscayl
|
||||
[3]: https://opensource.com/article/22/10/defining-open-source-ai
|
||||
[4]: https://appimage.github.io/Upscayl/
|
||||
[5]: https://github.com/upscayl/upscayl/blob/main/LICENSE
|
||||
[6]: https://opensource.com/sites/default/files/2023-03/upscayl-panel.webp
|
||||
[7]: https://github.com/xinntao/Real-ESRGAN
|
||||
[8]: https://upscale.wiki/wiki/Model_Database
|
||||
[9]: https://opensource.com/sites/default/files/2023-03/jurica-koletic-7YVZYZeITc8-unsplash.webp
|
Loading…
Reference in New Issue
Block a user