search

Artisense-ai / libartipy

libartipy is a set of tools for working with the 4Seasons dataset.
MIT License
14 stars 6 forks source link

Relative pose from imu to camera #2

Open WZG3661 opened 1 year ago

WZG3661 commented 1 year ago

Hi, I have some small questions about your dataset.

  1. In each sequence, there is a Transformations.txt which defines transformations between different coordinate frames. And the TS_cam_imu is the relative pose from imu to camera. What I want to know is that the pose is about the undistorted camera or the distorted camera. Because after the stereo rectified, the relative pose between camera and imu must be changed, right?
  2. What is the intrisics of imu, which inludes noise and random walk. I have looked up the sensor menu, but I only finded the noise. So, what values do you use?
Li123wen commented 1 year ago

嗨，我对你的数据集有一些小问题。

  1. 在每个序列中，都有一个定义不同坐标帧之间的转换。这是从imu到相机的相对姿势。我想知道的是，这个姿势是关于未失真的相机或失真的相机。因为立体声整流后，摄像头和imu之间的相对姿势必须改变，对吧？Transformations.txt``TS_cam_imu
  2. imu的内在是什么，它包含噪音和随机游走。我查找了传感器菜单，但我只找到了噪音。那么，您使用什么值？ hi!Recently, I also want to use the dataset of 4seasons datasets, but I cannot open the official interface to obtain the table. May I ask how you downloaded the dataset
QingCheng24 commented 1 week ago

To download, please check the download link. You can find all the related data here.

The TS_cam_imu is the relative pose of the undistorted camera

@NiclasZeller Do you have an answer for this

What is the intrisics of imu, which inludes noise and random walk. I have looked up the sensor menu, but I only finded the noise. So, what values do you use?

NiclasZeller commented 1 week ago

To download, please check the download link. You can find all the related data here.

The TS_cam_imu is the relative pose of the undistorted camera

@NiclasZeller Do you have an answer for this

What is the intrisics of imu, which inludes noise and random walk. I have looked up the sensor menu, but I only finded the noise. So, what values do you use?

We don't provide any noise model parameters with the dataset since we were not able to estimate the parameters accurately for our setup. However, some important information might be available in the datasheet of the imu. Furthermore, since DM-VIO achieves quite good results on our dataset, the parameters used might be quite reasonable.