Kinect v2 - >坐标框架校准？

问题描述

我的kinect v2传感器水平放置（不倾斜），高度为1.8米（直接位于电视屏幕上方），一个人坐在离传感器大约2.5-3米的位置。当这个人在
水平面上移动30厘米的距离时（在76厘米的高度，在传感器的方向上），估计的身体坐标是不正确的。虽然"腕关节"的实际高度保持不变，但估计该关节在运动终点高出10厘米左右，而不是起始点为
。这意味着该人必须向下移动他的手以产生与骨骼的水平运动，这严重干扰了实验任务的性能（为中风患者和帕金森病患者设计）。
有没有办法解决这个问题？

在我寻找这个问题的解决方案时，我偶然发现了一个'kinect for xbox '定义楼层的标准程序。是否有类似（或其他）的可能性来校准kinect v2 for windows？

 

解决方案

My kinect v2 sensor is positioned horizontally (not tilted) at a height of 1.8 m (directly above the TV screen) and a person is sitting at a distance of approximately 2.5-3 m from the sensor. When the person moves his hand over a distance of 30 cm in the horizontal plane (at a height of 76 cm, in direction of the sensor), the estimated body coordinates are incorrect. While the actual height of the 'wrist joint' kept constant, this joint is estimated ca 10 cm higher at the end point of the movement than at the start point. This means that the person has to move his hands downwards in order to generate a horizontal movement with the skeleton, which seriously interferes with performance of an experimental task (designed for stroke patients and Parkinson patients). Is there a way to solve this issue?

In my search for a solution to this problem, I stumbled upon a 'kinect for xbox' standard procedure to define the floor. Is there a similar (or other) possibility to calibrate kinect v2 for windows?

 

解决方案

这篇关于Kinect v2 - >坐标框架校准？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！