FOTS_TF icon indicating copy to clipboard operation
FOTS_TF copied to clipboard

about box_widths in data_utils.py

Open LXYTSOS opened this issue 6 years ago • 1 comments

I don't understand why did you divide the text_polyses by 4: x1, y1, x2, y2, x3, y3, x4, y4 = text_polyses[i] / 4, and what this line of code doing: width_box = math.ceil(8 * box_w / box_h), what's the meaning of 8 * box_w / box_h

LXYTSOS avatar Jan 16 '20 01:01 LXYTSOS

  1. Dividing by 4 is because of the size of feature map to apply RoI Rotate is a quarter of the input images
  2. The feature generated by RoI Rotate is with the size of (8 * ?), so 8 * box_w / box_h is to compute the width of RoI in a keep-ratio mode

Pay20Y avatar Jan 17 '20 11:01 Pay20Y