Fix text detection mapping to original image

MissouriMRR / IARC-2020

Missouri S&T's Multirotor Robot Design Team's 2020 International Aerial Robotics Competition code

MIT License

9 stars 3 forks source link

Closed fallscameron01 closed 3 years ago

fallscameron01 commented 3 years ago

Text detection currently does not accurately map detected text location on sliced section of image to the original image.

[x] After text is detected on sliced_rotated_image, the coordinates should be mapped back to the original image.
[x] Ensure the coordinates of the bounding boxes are correct when placed on the original image