That PS feature is not high dynamic range anything, which would require multiple images of the scene acquired with different exposure settings. The PS feature is a well tuned local contrast enhancement algorithm, for which there are plenty of google results.
The Photoshop plug in doesn't actually have a special algorithm. It just combines a number of different filters to create an image that displays similar characteristics with HDR images.
You can do the same thing with GIMP, following this Instructable. You might also play with a light edge-glow effect as well.
You may start with implementing "Unsharp Mask" based on Gaussian Blur. Pseudo HDR effects could be achieved by applying unsharp mask with large radius.
Combining it with "Shadow / Highlight" filter yields something similar to what you linked to.
发布评论
评论（3）
PS 功能并不是高动态范围，这需要使用不同曝光设置获取场景的多个图像。 PS 功能是一种经过良好调整的局部对比度增强算法，对此有大量的 google 结果。
以下是关于如何使用任何 C++ 图像框架执行此操作的可靠教程：http://www. Cromwell-intl.com/3d/histogram/
That PS feature is not high dynamic range anything, which would require multiple images of the scene acquired with different exposure settings. The PS feature is a well tuned local contrast enhancement algorithm, for which there are plenty of google results.
Here is a solid tutorial on how to do this with any C++ image framework: http://www.cromwell-intl.com/3d/histogram/
Photoshop 插件实际上没有特殊的算法。它只是结合了许多不同的滤镜来创建显示与 HDR 图像相似特征的图像。
您可以使用 GIMP 执行相同的操作，遵循此 可指导。您也可以使用轻微的边缘发光效果。
The Photoshop plug in doesn't actually have a special algorithm. It just combines a number of different filters to create an image that displays similar characteristics with HDR images.
You can do the same thing with GIMP, following this Instructable. You might also play with a light edge-glow effect as well.
您可以从实现基于高斯模糊的“USM 锐化”开始。
通过应用大半径的不锐化掩模可以实现伪 HDR 效果。
将其与“阴影/高光”过滤器结合起来会产生与您链接到的内容类似的效果。
You may start with implementing "Unsharp Mask" based on Gaussian Blur.
Pseudo HDR effects could be achieved by applying unsharp mask with large radius.
Combining it with "Shadow / Highlight" filter yields something similar to what you linked to.