ADM-201 dump PMP dumps pdf SSCP exam materials CBAP exam sample questions

图像的时空合成 – 译学馆
未登陆,请登陆后再发表信息
最新评论 (0)
播放视频

图像的时空合成

Space-Time Video Completion | Two Minute Papers

亲爱的学霸们大家好 这里是由Károly Zsolnai-Fehér带来的两分钟论文
Dear Fellow Scholars, this is Two Minute Papers with Károly Zsolnai-Fehér.
今天我们讲的算法能够填补图像中的缺失部分
Today we’re going to talk about an algorithm that is capable of filling holes
包括空间域和时间域上的缺失
in space and time.
传统的图像修复 或者说图像补洞算法
Classical image inpainting, or in other words, filling holes in images is something that
我们在之前的节目中说过 这个领域已有一些
we’ve explored in earlier episodes – there are several incredible techniques
很强的技术
to take care of that.
然而 本文将这种技术扩展到了视频上
But with this piece of work, it is possible to generalize such a technique for video,
不光可以给单幅图像补洞 还可以用于图像序列
and fill holes for not only one image but a series of images.
像是将视频中的雨伞去掉并创建一个无遮挡的沙滩视野
Like removing an umbrella to create an unoccluded view to the beach.
或者把我们慢跑视频中这个挥手的人去掉
Or removing a waving person from a video of us jogging.
这些结果真是太棒了 尽管这篇文章是很久很久以前发表的
These results are truly incredible, and even though this method was published long long
它仍然被许多计算机图像学应用者所尊敬
ago, it still enjoys a great deal of reverence among computer graphics practitioners.
不仅如此 这个算法还是Adobe Photoshop CS5中
Not only that, but this algorithm also serves the basis of the awesome content aware fill
引入的内容感知的图像填充技术的基础
feature introduced in Adobe Photoshop CS5.
在这个问题的构想中 必须保证我们的解决办法具有时空 一致性
In this problem formulation, we have to make sure that our solution has spatio-temporal consistency.
这到底意味着什么呢?
What does this mean exactly?
这意味着空洞可能存在与于空间和时间上
This means that holes can exist through space and time.
所以视频里可能有好几帧图像都丢失了
So multiple frames of a video may be missing,
或者我们希望删除的部分 不仅仅只针对某个画面 而是涉及整个视频
or there may be regions that we wish to cut out not only for one image, but for the entirety of the video.
补入的部分 从单幅画面来看 必须与其周围环境协调
The filled in regions have to be consistent with their surroundings if they are looked
而且必须在整个时域上保持前后一致
at as an image, but there also has to be a consistency across the time domain, otherwise
否则 我们会看到让人讨厌的闪烁效果
we would see a disturbing flickering effect in the results.
这确实是个很难的问题 因为对解法有更多的限制
It is a really challenging problem indeed because there are more constraints that we have to adhere to.
但是 我们发现它其实又更为简单
However, a key observation is that it is also easier
因为 作为回报 我们可以从视频的上一帧
because in return, we have access to more information
和下一帧中得到更多的信息
that comes from the previous and the next frames in the video.
比如 在这里 你可以看到通过删除一大片讨厌的人工痕迹
For instance, here, you can see an example of retouching old footage
来修饰原视频的例子
by removing a huge pesky artifact.
我们很清楚地知道 查理•卓别林应该在画面正中
And clearly we know the fact that Charlie Chaplin is supposed to be in the middle of
那仅仅是因为我们从上一帧和下一帧得到了这个信息
the image only because we have this information from the previous and next frames.
之所以能做到这些是因为我们找到了一个最优化的算法
All this is achieved by an optimization algorithm that takes into consideration that consistency
这个算法将时空一致性同时考虑进来
has to be enforced through the spatial and the time domain at the same time.
它也能够用来填补视频中完全缺失的画面
It can also be used to fill in completely missing frames of a video.
或者 它也能通过一个图像稳定算法 来帮助填补
Or, it also helps where we have parts of an image missing after being removed
那些被删除的部分
by an image stabilizer algorithm.
视频编辑人员总是在做这些 所以这一项恢复技术超级有用
Video editors do this all the time, so such a restoration technique is super useful.
这项技术的源程序可以在网上找到 我把链接贴在视频介绍里了
The source code of this technique is available, I’ve put a link to it in the video description.
感谢收看和支持 下次再见!
Thanks for watching and for your generous support, and I’ll see you next time!

发表评论

译制信息
视频概述

介绍了一种最优算法可以从空间和时间上来填补视频中的缺失部分,这种算法将时空一致性同时考虑进来。

听录译者

收集自网络

翻译译者

豆子

审核员

知易行难

视频来源

https://www.youtube.com/watch?v=YWK-bnyXvbg

相关推荐