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

preintegration accuracy #3

Open
cumtchenchang opened this issue Dec 17, 2023 · 2 comments
Open

preintegration accuracy #3

cumtchenchang opened this issue Dec 17, 2023 · 2 comments

Comments

@cumtchenchang
Copy link

作者你好,看到你的代码中对预积分进行了修改
0924389b3841c6b47008d3be533a524
不过我在自己的代码中进行了精度验证,没有正向效果。
还有想问一下有没有吃过CPI的形式:https://github.com/rpng/cpi。
祝好

@cumtchenchang
Copy link
Author

OKIVS对误差施加的是右扰动,对变量施加的是左扰动

@JzHuai0108
Copy link
Owner

JzHuai0108 commented Dec 18, 2023

OKVIS误差的定义是相对世界系的误差,我也采用了这用误差定义,这种定义可以看做是right invariant error的简化版。误差的定位是\f$ x = \delta x + \hat{x}\f$, \f$ R_{WB} = \exp(\delta\theta) \hat{R}_{WB}\f$, 其中\f$\hat{.}\f$表示估计量。
“OKIVS对误差施加的是右扰动” 不太理解这是啥意思哈。“对变量施加的是左扰动“ 我上面的定义应该是这样的吧。

关于截图中代码的修改,我推导发现preintegration factor covariance不太严格,所以做了一下的更改。原来的实现计算的factor covariance会偏大,参考这个文档
我代码验证了一下,修改后算出来的covariance和传统的机械编排以及gtsam算出的covariance很接近了。
关于正向改进,我做IMU组合实验的经验是,给IMU的噪声x (0.5 - 2)的系数,实验效果应该相差不大。所以我上面的改动也不太容易看出来哈。

我吃过CPI形式的预积分。Pardon me Dr Eckenhoff for not positive remarks.
个人感觉closed form integration是做了一些简化假设才实现的closed-form,对精度改进我感觉应该微弱。实际中,最流行的代码是最容易抄的代码,而不是复杂的精度好一些的,:)。

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