qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-06-10 20:59:54 +02:00
Code Issues Projects Releases Wiki Activity

few minor fixes

Browse source
This commit is contained in:
cneud 2025-03-26 00:27:06 +01:00
parent b4d168cae3
commit 62e8d78e73
3 changed files with 19 additions and 41 deletions
Download patch file Download diff file Expand all files Collapse all files

9
src/eynollah/sbb_binarize.py
View file

@ -1,27 +1,22 @@
""" """
Tool to load model and binarize a given image. Load model and binarize a given image.
""" """
import os
import sys import sys
from glob import glob from glob import glob
from os import environ, devnull from os import environ, devnull
from os.path import join from os.path import join
from warnings import catch_warnings, simplefilter from warnings import catch_warnings, simplefilter
import os
import numpy as np import numpy as np
from PIL import Image from PIL import Image
import cv2 import cv2
environ['TF_CPP_MIN_LOG_LEVEL'] = '3' environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
stderr = sys.stderr stderr = sys.stderr
sys.stderr = open(devnull, 'w') sys.stderr = open(devnull, 'w')
import tensorflow as tf import tensorflow as tf
from tensorflow.keras.models import load_model from tensorflow.keras.models import load_model
from tensorflow.python.keras import backend as tensorflow_backend from tensorflow.python.keras import backend as tensorflow_backend
sys.stderr = stderr sys.stderr = stderr
import logging import logging

45
src/eynollah/utils/drop_capitals.py
View file

@ -112,15 +112,12 @@ def adhere_drop_capital_region_into_corresponding_textline(
# print(arg_min) # print(arg_min)
cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min]) cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min])
cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 0] # +all_box_coord[int(region_final)][2]
0] # +all_box_coord[int(region_final)][2] cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 1] # +all_box_coord[int(region_final)][0]
cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0,
1] # +all_box_coord[int(region_final)][0]
img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3)) img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3))
img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255)) img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255))
img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], color=(255, 255, 255))
color=(255, 255, 255))
img_textlines = img_textlines.astype(np.uint8) img_textlines = img_textlines.astype(np.uint8)
@ -180,15 +177,12 @@ def adhere_drop_capital_region_into_corresponding_textline(
# print(arg_min) # print(arg_min)
cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min]) cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min])
cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 0] # +all_box_coord[int(region_final)][2]
0] # +all_box_coord[int(region_final)][2] cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 1] # +all_box_coord[int(region_final)][0]
cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0,
1] # +all_box_coord[int(region_final)][0]
img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3)) img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3))
img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255)) img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255))
img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], color=(255, 255, 255))
color=(255, 255, 255))
img_textlines = img_textlines.astype(np.uint8) img_textlines = img_textlines.astype(np.uint8)
@ -238,15 +232,12 @@ def adhere_drop_capital_region_into_corresponding_textline(
# print(arg_min) # print(arg_min)
cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min]) cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min])
cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, cnt_nearest[:, 0, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 0] # +all_box_coord[int(region_final)][2]
0] # +all_box_coord[int(region_final)][2] cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0, 1] # +all_box_coord[int(region_final)][0]
cnt_nearest[:, 0, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 0,
1] # +all_box_coord[int(region_final)][0]
img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3)) img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3))
img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255)) img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255))
img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], color=(255, 255, 255))
color=(255, 255, 255))
img_textlines = img_textlines.astype(np.uint8) img_textlines = img_textlines.astype(np.uint8)
contours_combined = return_contours_of_interested_region(img_textlines, 255, 0) contours_combined = return_contours_of_interested_region(img_textlines, 255, 0)
@ -356,15 +347,12 @@ def adhere_drop_capital_region_into_corresponding_textline(
# print(arg_min) # print(arg_min)
cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min]) cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min])
cnt_nearest[:, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0] + \ cnt_nearest[:, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0] + all_box_coord[int(region_final)][2]
all_box_coord[int(region_final)][2] cnt_nearest[:, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 1] + all_box_coord[int(region_final)][0]
cnt_nearest[:, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 1] + \
all_box_coord[int(region_final)][0]
img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3)) img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3))
img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255)) img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255))
img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], color=(255, 255, 255))
color=(255, 255, 255))
img_textlines = img_textlines.astype(np.uint8) img_textlines = img_textlines.astype(np.uint8)
contours_combined = return_contours_of_interested_region(img_textlines, 255, 0) contours_combined = return_contours_of_interested_region(img_textlines, 255, 0)
@ -424,15 +412,12 @@ def adhere_drop_capital_region_into_corresponding_textline(
# print(arg_min) # print(arg_min)
cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min]) cnt_nearest = np.copy(all_found_textline_polygons[int(region_final)][arg_min])
cnt_nearest[:, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0] + \ cnt_nearest[:, 0] = all_found_textline_polygons[int(region_final)][arg_min][:, 0] + all_box_coord[int(region_final)][2]
all_box_coord[int(region_final)][2] cnt_nearest[:, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 1] + all_box_coord[int(region_final)][0]
cnt_nearest[:, 1] = all_found_textline_polygons[int(region_final)][arg_min][:, 1] + \
all_box_coord[int(region_final)][0]
img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3)) img_textlines = np.zeros((text_regions_p.shape[0], text_regions_p.shape[1], 3))
img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255)) img_textlines = cv2.fillPoly(img_textlines, pts=[cnt_nearest], color=(255, 255, 255))
img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], img_textlines = cv2.fillPoly(img_textlines, pts=[polygons_of_drop_capitals[i_drop]], color=(255, 255, 255))
color=(255, 255, 255))
img_textlines = img_textlines.astype(np.uint8) img_textlines = img_textlines.astype(np.uint8)
contours_combined = return_contours_of_interested_region(img_textlines, 255, 0) contours_combined = return_contours_of_interested_region(img_textlines, 255, 0)

6
src/eynollah/utils/marginals.py
View file

@ -126,8 +126,7 @@ def get_marginals(text_with_lines, text_regions, num_col, slope_deskew, light_ve
polygons_of_marginals = return_contours_of_interested_region(text_regions, pixel_img, min_area_text) polygons_of_marginals = return_contours_of_interested_region(text_regions, pixel_img, min_area_text)
cx_text_only, cy_text_only, x_min_text_only, x_max_text_only, y_min_text_only, y_max_text_only, y_cor_x_min_main = find_new_features_of_contours( (cx_text_only, cy_text_only, x_min_text_only, x_max_text_only, y_min_text_only, y_max_text_only, y_cor_x_min_main) = (find_new_features_of_contours(polygons_of_marginals))
polygons_of_marginals)
text_regions[(text_regions[:, :] == 4)] = 1 text_regions[(text_regions[:, :] == 4)] = 1
@ -155,8 +154,7 @@ def get_marginals(text_with_lines, text_regions, num_col, slope_deskew, light_ve
polygons_of_marginals = return_contours_of_interested_region(text_regions, pixel_img, min_area_text) polygons_of_marginals = return_contours_of_interested_region(text_regions, pixel_img, min_area_text)
cx_text_only, cy_text_only, x_min_text_only, x_max_text_only, y_min_text_only, y_max_text_only, y_cor_x_min_main = find_new_features_of_contours( (cx_text_only, cy_text_only, x_min_text_only, x_max_text_only, y_min_text_only, y_max_text_only, y_cor_x_min_main) = (find_new_features_of_contours(polygons_of_marginals))
polygons_of_marginals)
text_regions[(text_regions[:, :] == 4)] = 1 text_regions[(text_regions[:, :] == 4)] = 1