search

rpng / R-VIO

Robocentric Visual-Inertial Odometry
https://journals.sagepub.com/doi/10.1177/0278364919853361
GNU General Public License v3.0
753 stars 174 forks source link

Does Gaussian thresholding and Box blurring help? #11

Closed JiatianWu closed 5 years ago

JiatianWu commented 5 years ago

I noticed that you mentioned using Guassian thresholding and box blurring on image before doing the KLT track, but I did not find implementation in your code. I wonder that do these two tricks help with image blur/dark/ varying light problems?

huaizheng commented 5 years ago

To present a complete VIO system, in the current front end I adopt standard histogram equalization to preprocess the image, which is commonly used to deal with the illumination change. Also, in my paper I instead used the above combination to process the image and help the visual tracking. As I have modularized the R-VIO codebase, you could also try different front end with our proposed back end (robocentric estimator) to see whether it improves the performance, which I think would be interesting.