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

Initialization not correct #4

Open
rising-turtle opened this issue Jun 16, 2018 · 1 comment
Open

Initialization not correct #4

rising-turtle opened this issue Jun 16, 2018 · 1 comment

Comments

@rising-turtle
Copy link

Hi, all
I use the default config file to run the VINS-Sample, however, the result of the initialization lead to large scale error. I think the default config file provide pretty good initial estimate for the IMU-Camera transformation, which should not be the main reason that results in the scale error.

To run the code, I follow the instructions in the README file of the MYNT-EYE-VINS-Sample. Is there something I missed?

Any suggestions will be greatly appreciated!

Thanks.

@jiziye
Copy link

jiziye commented Oct 29, 2019

Hi,

  1. Please make sure the camera uses the latest version of the SDK and the latest firmware.

  2. Has the camera been dropped or bumped? What is the depth map effect? If the depth effect is poor, the parameters are no longer adapted and need to be recalibrated.

  3. Whether the imu data is normal, you can check the data of the iMU in the static state. Need to ensure that the imu data is normal.

  4. Before running vins, you need to set up the camera. Facing the place with many feature points, after the vins starts, slowly move the camera to initialize. The camera can't lie down and initialize, it will fly. Also, when running vins, turn off the IR and don't use IR. In addition, ceres need to use 1.14.0, eigen needs to use version 3.3.4, the wrong version will also cause vins to fly.

Best.

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