|
|
|
@ -195,6 +195,8 @@ class Eynollah:
|
|
|
|
|
logger=None,
|
|
|
|
|
pcgts=None,
|
|
|
|
|
):
|
|
|
|
|
if skip_layout_and_reading_order:
|
|
|
|
|
textline_light = True
|
|
|
|
|
self.light_version = light_version
|
|
|
|
|
if not dir_in:
|
|
|
|
|
if image_pil:
|
|
|
|
@ -1512,7 +1514,7 @@ class Eynollah:
|
|
|
|
|
textlines_ins = [polygons_of_textlines[ind] for ind in indexes_in]
|
|
|
|
|
|
|
|
|
|
all_found_textline_polygons.append(textlines_ins)
|
|
|
|
|
slopes.append(0)
|
|
|
|
|
slopes.append(slope_deskew)
|
|
|
|
|
|
|
|
|
|
_, crop_coor = crop_image_inside_box(boxes[index],image_page_rotated)
|
|
|
|
|
|
|
|
|
@ -1527,11 +1529,8 @@ class Eynollah:
|
|
|
|
|
results = self.executor.map(partial(do_work_of_slopes_new_light,
|
|
|
|
|
textline_mask_tot_ea=textline_mask_tot,
|
|
|
|
|
image_page_rotated=image_page_rotated,
|
|
|
|
|
slope_deskew=slope_deskew,
|
|
|
|
|
MAX_SLOPE=MAX_SLOPE,
|
|
|
|
|
KERNEL=KERNEL,
|
|
|
|
|
logger=self.logger,
|
|
|
|
|
plotter=self.plotter,),
|
|
|
|
|
slope_deskew=slope_deskew,textline_light=self.textline_light,
|
|
|
|
|
logger=self.logger,),
|
|
|
|
|
boxes, contours, contours_par, range(len(contours_par)))
|
|
|
|
|
#textline_polygons, boxes, text_regions, text_regions_par, box_coord, index_text_con, slopes = zip(*results)
|
|
|
|
|
self.logger.debug("exit get_slopes_and_deskew_new_light")
|
|
|
|
@ -4245,7 +4244,7 @@ class Eynollah:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
def filter_contours_without_textline_inside(self,contours,text_con_org, contours_textline):
|
|
|
|
|
def filter_contours_without_textline_inside(self,contours,text_con_org, contours_textline, contours_only_text_parent_d_ordered):
|
|
|
|
|
|
|
|
|
|
###contours_txtline_of_all_textregions = []
|
|
|
|
|
|
|
|
|
@ -4282,8 +4281,9 @@ class Eynollah:
|
|
|
|
|
contours.pop(ind_u_a_trs)
|
|
|
|
|
contours_textline.pop(ind_u_a_trs)
|
|
|
|
|
text_con_org.pop(ind_u_a_trs)
|
|
|
|
|
contours_only_text_parent_d_ordered.pop(ind_u_a_trs)
|
|
|
|
|
|
|
|
|
|
return contours, text_con_org, contours_textline
|
|
|
|
|
return contours, text_con_org, contours_textline, contours_only_text_parent_d_ordered, np.array(range(len(contours)))
|
|
|
|
|
|
|
|
|
|
def dilate_textlines(self,all_found_textline_polygons):
|
|
|
|
|
for j in range(len(all_found_textline_polygons)):
|
|
|
|
@ -4487,12 +4487,9 @@ class Eynollah:
|
|
|
|
|
continue
|
|
|
|
|
else:
|
|
|
|
|
return pcgts
|
|
|
|
|
|
|
|
|
|
if self.skip_layout_and_reading_order:
|
|
|
|
|
img_res, is_image_enhanced, num_col_classifier, num_column_is_classified = self.run_enhancement(self.light_version)
|
|
|
|
|
self.logger.info("Enhancing took %.1fs ", time.time() - t0)
|
|
|
|
|
#print("text region early -1 in %.1fs", time.time() - t0)
|
|
|
|
|
t1 = time.time()
|
|
|
|
|
if self.skip_layout_and_reading_order:
|
|
|
|
|
_ ,_, _, textline_mask_tot_ea, img_bin_light = self.get_regions_light_v(img_res, is_image_enhanced, num_col_classifier,
|
|
|
|
|
skip_layout_and_reading_order=self.skip_layout_and_reading_order)
|
|
|
|
|
|
|
|
|
@ -4522,16 +4519,20 @@ class Eynollah:
|
|
|
|
|
polygons_lines_xml = []
|
|
|
|
|
contours_tables = []
|
|
|
|
|
ocr_all_textlines = None
|
|
|
|
|
|
|
|
|
|
pcgts = self.writer.build_pagexml_no_full_layout(cont_page, page_coord, order_text_new, id_of_texts_tot,
|
|
|
|
|
all_found_textline_polygons, page_coord, polygons_of_images, polygons_of_marginals,
|
|
|
|
|
all_found_textline_polygons_marginals, all_box_coord_marginals, slopes, slopes_marginals,
|
|
|
|
|
cont_page, polygons_lines_xml, contours_tables, ocr_all_textlines)
|
|
|
|
|
if self.dir_in:
|
|
|
|
|
self.writer.write_pagexml(pcgts)
|
|
|
|
|
continue
|
|
|
|
|
else:
|
|
|
|
|
return pcgts
|
|
|
|
|
|
|
|
|
|
if not self.extract_only_images and not self.skip_layout_and_reading_order:
|
|
|
|
|
img_res, is_image_enhanced, num_col_classifier, num_column_is_classified = self.run_enhancement(self.light_version)
|
|
|
|
|
self.logger.info("Enhancing took %.1fs ", time.time() - t0)
|
|
|
|
|
#print("text region early -1 in %.1fs", time.time() - t0)
|
|
|
|
|
t1 = time.time()
|
|
|
|
|
if self.light_version:
|
|
|
|
|
text_regions_p_1 ,erosion_hurts, polygons_lines_xml, textline_mask_tot_ea, img_bin_light = self.get_regions_light_v(img_res, is_image_enhanced, num_col_classifier)
|
|
|
|
|
#print("text region early -2 in %.1fs", time.time() - t0)
|
|
|
|
@ -4770,7 +4771,7 @@ class Eynollah:
|
|
|
|
|
all_found_textline_polygons, boxes_text, txt_con_org, contours_only_text_parent, all_box_coord, index_by_text_par_con, slopes = \
|
|
|
|
|
self.get_slopes_and_deskew_new_light2(txt_con_org, contours_only_text_parent, textline_mask_tot_ea_org, image_page_rotated, boxes_text, slope_deskew)
|
|
|
|
|
all_found_textline_polygons_marginals, boxes_marginals, _, polygons_of_marginals, all_box_coord_marginals, _, slopes_marginals = \
|
|
|
|
|
self.get_slopes_and_deskew_new_light(polygons_of_marginals, polygons_of_marginals, textline_mask_tot_ea_org, image_page_rotated, boxes_marginals, slope_deskew)
|
|
|
|
|
self.get_slopes_and_deskew_new_light2(polygons_of_marginals, polygons_of_marginals, textline_mask_tot_ea_org, image_page_rotated, boxes_marginals, slope_deskew)
|
|
|
|
|
|
|
|
|
|
#slopes, all_found_textline_polygons, boxes_text, txt_con_org, contours_only_text_parent, index_by_text_par_con = \
|
|
|
|
|
# self.delete_regions_without_textlines(slopes, all_found_textline_polygons, boxes_text, txt_con_org, contours_only_text_parent, index_by_text_par_con)
|
|
|
|
@ -4783,7 +4784,7 @@ class Eynollah:
|
|
|
|
|
all_found_textline_polygons = self.filter_contours_inside_a_bigger_one(all_found_textline_polygons, textline_mask_tot_ea_org, type_contour="textline")
|
|
|
|
|
all_found_textline_polygons_marginals = self.dilate_textregions_contours_textline_version(all_found_textline_polygons_marginals)
|
|
|
|
|
|
|
|
|
|
contours_only_text_parent, txt_con_org, all_found_textline_polygons = self.filter_contours_without_textline_inside(contours_only_text_parent,txt_con_org, all_found_textline_polygons)
|
|
|
|
|
contours_only_text_parent, txt_con_org, all_found_textline_polygons, contours_only_text_parent_d_ordered, index_by_text_par_con = self.filter_contours_without_textline_inside(contours_only_text_parent,txt_con_org, all_found_textline_polygons, contours_only_text_parent_d_ordered)
|
|
|
|
|
|
|
|
|
|
else:
|
|
|
|
|
textline_mask_tot_ea = cv2.erode(textline_mask_tot_ea, kernel=KERNEL, iterations=1)
|
|
|
|
@ -4885,12 +4886,12 @@ class Eynollah:
|
|
|
|
|
ocr_all_textlines = []
|
|
|
|
|
else:
|
|
|
|
|
ocr_all_textlines = None
|
|
|
|
|
|
|
|
|
|
pcgts = self.writer.build_pagexml_full_layout(contours_only_text_parent, contours_only_text_parent_h, page_coord, order_text_new, id_of_texts_tot,
|
|
|
|
|
all_found_textline_polygons, all_found_textline_polygons_h, all_box_coord, all_box_coord_h, polygons_of_images, contours_tables, polygons_of_drop_capitals, polygons_of_marginals,
|
|
|
|
|
all_found_textline_polygons_marginals, all_box_coord_marginals, slopes, slopes_h, slopes_marginals,
|
|
|
|
|
cont_page, polygons_lines_xml, ocr_all_textlines)
|
|
|
|
|
self.logger.info("Job done in %.1fs", time.time() - t0)
|
|
|
|
|
print("Job done in %.1fs", time.time() - t0)
|
|
|
|
|
if self.dir_in:
|
|
|
|
|
self.writer.write_pagexml(pcgts)
|
|
|
|
|
continue
|
|
|
|
|