简体   繁体   English

Kinect v2 SDK 2.0如何将多个骨架主题保存到文件

[英]Kinect v2 SDK 2.0 How to Save Multiple Skeleton Subjects to File

My question is pretty straightforward. 我的问题很简单。 I can save all streams of Kinect v2 (and Kinect v1) synchronized. 我可以同步保存所有Kinect v2 (and Kinect v1)流。 If I also have skeleton for one person, skeleton information is also saved. 如果我也有一个人的骨架,也会保存骨架信息。

My issue now is that I want to save the same synchronized information (color, depth, infrared, bodyindex and skeleton), but for the maximum of 6 people. 我现在的问题是我想保存相同的同步信息(颜色,深度,红外线,身体指数和骨骼),但最多可以保存6个人。 I think this can be accomplished by one of two ways: 我认为这可以通过以下两种方式之一来完成:

Solution 1) - create one skeleton file for each of the 6 people, using the ID as reference for the filename, for instance. 解决方案1) - 为6个人中的每个人创建一个骨架文件,例如使用ID作为文件名的参考。

Solution 2) - saved all this information in one single file, as I do for the remaining streams, using again an ID before each new person. 解决方案2) - 将所有这些信息保存在一个文件中,就像我对其余流一样,在每个新人之前再次使用ID。

Besides, anyone has any clue on how to "lock" each person ID using Kinect SDK 2.0 ? 此外,任何人都有任何关于如何使用Kinect SDK 2.0 “锁定”每个人ID的线索?

我认为最好为每个人创建单个文件。然后你可以很容易地改变内容。如果你在一个文件中添加所有细节,它将使代码变得复杂。

Personally, I would save the data under a single INI or XML file using the persons ID to create the section they are referenced by. 就个人而言,我会使用人员ID将数据保存在单个INI或XML文件下,以创建他们所引用的部分。 This will help reduce the complexity of your code if you use an online library. 如果您使用在线库,这将有助于降低代码的复杂性。 This will allow for greater extensibility and offer you the ability to edit two profiles at once without having to open multiple files. 这将允许更大的可扩展性,并使您能够一次编辑两个配置文件,而无需打开多个文件。

As for 'locking' the user to an ID, that isn't something the Kinect itself is capable of. 至于将用户“锁定”到ID,这不是Kinect本身能够做到的。 In order to do that you would have to create an algorithm that assigns an ID based on information about the user. 为此,您必须创建一个算法,根据有关用户的信息分配ID。 To create an ID with staying power you need to make your program reliably be able to assign a unique signifier based on something like facial recognition, body size, name and/or distinguishing characteristics. 要创建具有持久力的ID,您需要使您的程序能够可靠地分配基于面部识别,身体大小,名称和/或区别特征之类的独特能指。 The ID would have to be system independent, meaning that the same ID would be ascribed to that person regardless of if the computer running the program has seen them before. ID必须与系统无关,这意味着无论运行程序的计算机是否曾经看过它,都会将相同的ID归于该人。 This would allow you to not only keep track of someone who leaves the screen, but between systems. 这样,您不仅可以跟踪离开屏幕的人,还可以跟踪系统之间的人。 This, however, is very difficult to coordinate and there is very little middle ground in making a system able to know that the person walking back into frame is the same. 然而,这很难协调,并且在使系统能够知道走回框架的人是相同的情况下几乎没有中间立场。 The Kinect DOES do this, but it uses xbox accounts as its unique identifier, something that the SDK does not allow you to utilize. Kinect DOES会这样做,但它使用xbox帐户作为其唯一标识符,这是SDK不允许您使用的。 (TL;DR You would have to make your own version of xbox accounts and your own identifying system for those accounts using the SDK) (TL; DR您必须使用SDK为这些帐户创建自己的xbox帐户版本和您自己的识别系统)

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM