qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-12-01 08:44:13 +01:00
Code Issues Projects Releases Wiki Activity

return_boxes_of_images_by_order_of_reading_new: improve

Browse source 
- when analysing regions spanning across columns,
  disregard tiny regions (smaller than half the median size)
- if a region spans across columns just by a tiny fraction,
  and therefore is not good enough for a multi-col separator,
  then it should also not be good enough for a multi-col box
  maker
This commit is contained in:
Robert Sachunsky 2025-11-28 17:58:44 +01:00
parent b71bb80e3a
commit 5abf0c1097
1 changed files with 8 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

12
src/eynollah/utils/__init__.py
View file

@ -1720,6 +1720,7 @@ def return_boxes_of_images_by_order_of_reading_new(
# analyse connected components of regions to gain additional separators # analyse connected components of regions to gain additional separators
# and prepare a map for cross-column boxes # and prepare a map for cross-column boxes
ccounts = np.bincount(ccomps[top: bot].flatten()) ccounts = np.bincount(ccomps[top: bot].flatten())
ccounts_median = np.median(ccounts)
col_ccounts = np.stack([np.bincount(ccomps[top: bot, left: right].flatten(), col_ccounts = np.stack([np.bincount(ccomps[top: bot, left: right].flatten(),
minlength=ccounts.size) minlength=ccounts.size)
for left, right in pairwise(peaks_neg_tot)]) for left, right in pairwise(peaks_neg_tot)])
@ -1727,6 +1728,9 @@ def return_boxes_of_images_by_order_of_reading_new(
for label, label_count in enumerate(ccounts): for label, label_count in enumerate(ccounts):
if not label: if not label:
continue continue
# ignore small labels for the purpose of finding multicol seps
if label_count < 0.5 * ccounts_median:
continue
label_left, label_top, label_width, label_height, label_area = cstats[label] label_left, label_top, label_width, label_height, label_area = cstats[label]
# if label_count < 0.9 * label_area: # if label_count < 0.9 * label_area:
# # mostly not in this part of the page # # mostly not in this part of the page
@ -1738,15 +1742,15 @@ def return_boxes_of_images_by_order_of_reading_new(
label_bot = label_top + label_height label_bot = label_top + label_height
label_start = np.flatnonzero(peaks_neg_tot > label_left)[0] - 1 label_start = np.flatnonzero(peaks_neg_tot > label_left)[0] - 1
label_end = np.flatnonzero(peaks_neg_tot >= label_right)[0] label_end = np.flatnonzero(peaks_neg_tot >= label_right)[0]
if label_end - label_start < 2:
continue
if np.count_nonzero(col_ccounts[:, label] > 0.1 * label_count) < 2:
continue
# store as dict for multi-column boxes: # store as dict for multi-column boxes:
for start in range(label_start, label_end): for start in range(label_start, label_end):
labelcolmap.setdefault(start, list()).append( labelcolmap.setdefault(start, list()).append(
(label_end, label_top, label_bot, sum(col_ccounts[start: label_end, label]))) (label_end, label_top, label_bot, sum(col_ccounts[start: label_end, label])))
# make additional separators: # make additional separators:
if label_end - label_start < 2:
continue
if np.count_nonzero(col_ccounts[:, label] > 0.1 * label_count) < 2:
continue
x_min_hor_some = np.append(x_min_hor_some, [label_left] * 2) x_min_hor_some = np.append(x_min_hor_some, [label_left] * 2)
x_max_hor_some = np.append(x_max_hor_some, [label_right] * 2) x_max_hor_some = np.append(x_max_hor_some, [label_right] * 2)
y_min_hor_some = np.append(y_min_hor_some, [label_top - 2, label_bot]) y_min_hor_some = np.append(y_min_hor_some, [label_top - 2, label_bot])