记录Kinect Fusion实时重建 [英] Recording Kinect Fusion real time reconstruction

问题描述

是否可以将kinect融合重建作为视频提取？

Is it possible to extract kinect fusion reconstruction as a video?

假设我有一个非常强大的GPU，能够实现几乎实时的移动物体重建使用Kinect Fusion，我想知道是否有办法将重建作为3D帧的结合提取。 （我知道kinect studio
可以记录深度流，但这不是我的兴趣）。

Assuming i have a very powerfull GPU that would be capable of achieve almost real time reconstruction of a moving object with Kinect Fusion, i want to know if there is a way to extract the reconstruction as a conjunct of 3D frames. (i know kinect studio can record depth stream, but that isnt my interest).

推荐答案

获取重建的卷信息是一个导出函数，它将为您提供点/顶点信息。我不知道"实时"如何这将是从中生成视频。您在应用中看到的可视化，以图像形式提供，因此
同样适用于Kinect中的其他帧数据，将其保存为视频中可用的内容。

getting the reconstructed volume information is an export function which will provide you point/vertex information. I don't know how "real-time" that would be to generate a video from. The visualization you see in the app, is provided as an image, so the same applies to other frame data from Kinect when saving that as something usuable in your video.

Fusion设计用于静态场景，您应该尽量减少更改量。您可以进行快速测试，以查看重建检测和删除引入的对象所需的时间。

Fusion was designed to work with a static scene and you should minimize the amount of change. You can do a quick test to see how long the reconstruction takes to detect and remove objects introduced.

这篇关于记录Kinect Fusion实时重建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！