FAST_LIO

FAST_LIO copied to clipboard

Hi, thank you for the great work on FAST-LIO2.

While studying both the paper and the code, I noticed a discrepancy between Equation (14) in the paper and its corresponding implementation in the code here.

In the paper, Equation (14) includes the inverse of the Jacobian matrix (𝐽𝜅)−1 , but this term appears to be omitted in the code. Although the paper briefly explains the role of this Jacobian, I’m still struggling to grasp its intuitive meaning.

Additionally, on a more fundamental level, I find it difficult to understand why the update formula is not simply implemented as:

dx_new + K_h Could you please clarify:

1.Why is the inverse Jacobian omitted in the code?

2.What is the intuitive role of (𝐽𝜅)−1 in the iterated update?

3.Why is the update formula not expressed in the form dx_new + K_h?

Thank you in advance!