无效深度像素，kinect如何填充黑洞？ [英] invalid depth pixel, how do kinect fill the black hole?

问题描述

我发现许多微软研究已经做了很好的预处理。像这样：

I found that many microsoft research had already made a good preprocess. Like these:

在KinectFusion论文中，它提到了一种技术被称为"Phong-shaded renderings"的是进行这样的深度估计吗？

In KinectFusion paper, it mentions a technique called "Phong-shaded renderings", is that make such depth estimation?

我还发现了其他一些方法，例如

and I also found some other approach, such as

 "Anat Levin Dani，Dani Lischinski和Yair Weiss。使用优化着色。 ACM Transactions on Graphics，23：689-694,2004"

我想知道关于无效深度像素的官方解决方案是什么，是否有一些文档或SDK代码可以了解更多信息？ ：）

推荐答案

没有官方解决方案，如果有的话，它可能是SDK的一部分。有许多方案和技术，其中一个可能比其他方案更好。 由开发人员实施最适合他们的方案。记住
，不是每个人都同意速度与质量之间的相同权衡。

There isn't an official solution, if there was, it probably would be part of the SDK. There are many scenarios and techniques where one might work better than other. It is left to the developer to implement which works best for them. Keep in mind, not everyone can agree on the same trade-offs between speed vs. quality.

这篇关于无效深度像素，kinect如何填充黑洞？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！