Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

DARTset可视化 #13

Open
Divenirelrn opened this issue Oct 13, 2023 · 2 comments
Open

DARTset可视化 #13

Divenirelrn opened this issue Oct 13, 2023 · 2 comments

Comments

@Divenirelrn
Copy link

你好,感谢您开源的项目,我有几个问题不理解,想请教下:
(1)您开源的DARTset数据集的3D坐标是什么坐标系下的?
(2)在DARTset.py文件第95行中,为什么要对y、z坐标取反?
image
(3)在DARTset.py文件第97、98行中,为什么要做平移,是为了更好地可视化吗?
(4)我自己用Unity GUI工具做出来的图像与标签,用postprocess目录下的convert.py处理后,再用DARTset.py进行可视化,为什么3D坐标投影后与2D坐标对不上?

@tomguluson92
Copy link
Collaborator

hello,
(1) 3D坐标就是MANO原始的坐标系, 我很久没检查了.
image
(2) 取反是为了把Unity里面的坐标系搞成我们的坐标系.
(3) 是的, 这是为了渲染object-centric的结果
(4) 具体对不上的话可以放出来debug下?

@Divenirelrn
Copy link
Author

Divenirelrn commented Oct 16, 2023

好的,感谢回复,关于(4),我做了以下流程:
1.用官网下载的DART GUI输出一张图像及对应的标签文件:
GUI下载方式:
image
得到的图像与标签:
image
图像与标签文件可从以下链接下载:GoogleDrive或者百度网盘 (提取码:l4hw)
2.用https://github.com/DART2022/DART/blob/master/postprocess/convert.py将输出的标签转化为output.obj:
image
output.obj也可从1中链接下载
3.用https://github.com/DART2022/DART/blob/master/DARTset.py进行可视化:
(1)首先,读取output.obj的2d坐标joints_2d与3d坐标joints_3d:
image
image
(2)然后,利用joints_2d与joints_3d计算投影矩阵ortho_intr:
image
image
(3)再利用joints_3d与ortho_intr进行投影,得到投影后的2d坐标proj_2d:
image
image
(4)最后将joints_2d与proj_2d画在同一张图像上:
image
得到以下结果:
image
该图像中,joints_2d的坐标是准确的,但proj_2d不准确,joints_2d与proj_2d对不上,请问这是什么原因?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants