qurator-spk/eynollah
1
0
Fork 0
mirror of https://github.com/qurator-spk/eynollah.git synced 2025-08-29 03:49:54 +02:00
Code Issues Projects Releases Wiki Activity

enhancing marginal detection for light version

Browse source
This commit is contained in:
vahidrezanezhad 2025-05-18 02:48:05 +02:00
parent 0819730355
commit 7a34bbb493
3 changed files with 13 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

7
src/eynollah/eynollah.py
View file

@ -272,7 +272,7 @@ class Eynollah:
self.model_region_dir_p_ens = dir_models + "/eynollah-main-regions-ensembled_20210425"
self.model_region_dir_p_ens_light = dir_models + "/eynollah-main-regions_20220314"
self.model_region_dir_p_ens_light_only_images_extraction = dir_models + "/eynollah-main-regions_20231127_672_org_ens_11_13_16_17_18"
self.model_reading_order_dir = dir_models + "/model_ens_reading_order_machine_based"
self.model_reading_order_dir = dir_models + "/model_step_2500000_mb_ro"#"/model_ens_reading_order_machine_based"
#"/modelens_12sp_elay_0_3_4__3_6_n"
#"/modelens_earlylayout_12spaltige_2_3_5_6_7_8"
#"/modelens_early12_sp_2_3_5_6_7_8_9_10_12_14_15_16_18"
@ -1315,7 +1315,7 @@ class Eynollah:
seg_art[seg_art>0] =1
seg_line = label_p_pred[:,:,:,3]
seg_line[seg_line>0.5] =1#seg_line[seg_line>0.1] =1
seg_line[seg_line>0.3] =1#seg_line[seg_line>0.5] =1#seg_line[seg_line>0.1] =1
seg_line[seg_line<1] =0
##seg[seg_art==1]=4
@ -3667,7 +3667,6 @@ class Eynollah:
peaks_real, _ = find_peaks(sum_smoothed, height=0)
if len(peaks_real)>70:
print(len(peaks_real), 'len(peaks_real)')
peaks_real = peaks_real[(peaks_real<width2) & (peaks_real>width1)]
@ -5133,7 +5132,7 @@ class Eynollah_ocr:
self.b_s = int(batch_size)
else:
self.model_ocr_dir = dir_models + "/model_step_425000_ocr"#"/model_step_125000_ocr"#"/model_step_25000_ocr"#"/model_step_1050000_ocr"#"/model_0_ocr_cnnrnn"#"/model_23_ocr_cnnrnn"
self.model_ocr_dir = dir_models + "/model_step_600000_ocr"#"/model_step_125000_ocr"#"/model_step_25000_ocr"#"/model_step_1050000_ocr"#"/model_0_ocr_cnnrnn"#"/model_23_ocr_cnnrnn"
model_ocr = load_model(self.model_ocr_dir , compile=False)
self.prediction_model = tf.keras.models.Model(

13
src/eynollah/utils/marginals.py
View file

@ -10,7 +10,6 @@ def get_marginals(text_with_lines, text_regions, num_col, slope_deskew, light_ve
mask_marginals=np.zeros((text_with_lines.shape[0],text_with_lines.shape[1]))
mask_marginals=mask_marginals.astype(np.uint8)
text_with_lines=text_with_lines.astype(np.uint8)
##text_with_lines=cv2.erode(text_with_lines,self.kernel,iterations=3)
@ -26,9 +25,11 @@ def get_marginals(text_with_lines, text_regions, num_col, slope_deskew, light_ve
text_with_lines=resize_image(text_with_lines,int(text_with_lines.shape[0]*1.8),text_with_lines.shape[1])
text_with_lines=cv2.erode(text_with_lines,kernel,iterations=7)
text_with_lines=resize_image(text_with_lines,text_with_lines_eroded.shape[0],text_with_lines_eroded.shape[1])
if light_version:
text_with_lines=rotate_image(text_with_lines,-slope_deskew)
kernel_hor = np.ones((1, 5), dtype=np.uint8)
text_with_lines = cv2.erode(text_with_lines,kernel_hor,iterations=6)
text_with_lines_y=text_with_lines.sum(axis=0)
text_with_lines_y_eroded=text_with_lines_eroded.sum(axis=0)
@ -42,8 +43,10 @@ def get_marginals(text_with_lines, text_regions, num_col, slope_deskew, light_ve
elif thickness_along_y_percent>=30 and thickness_along_y_percent<50:
min_textline_thickness=20
else:
min_textline_thickness=40
if light_version:
min_textline_thickness=45
else:
min_textline_thickness=40
if thickness_along_y_percent>=14:

4
src/eynollah/utils/separate_lines.py
View file

@ -1466,7 +1466,7 @@ def return_deskew_slop(img_patch_org, sigma_des,n_tot_angles=100,
main_page=False, logger=None, plotter=None, map=map):
if main_page and plotter:
plotter.save_plot_of_textline_density(img_patch_org)
img_int=np.zeros((img_patch_org.shape[0],img_patch_org.shape[1]))
img_int[:,:]=img_patch_org[:,:]#img_patch_org[:,:,0]
@ -1487,7 +1487,7 @@ def return_deskew_slop(img_patch_org, sigma_des,n_tot_angles=100,
angles = np.linspace(angle - 22.5, angle + 22.5, n_tot_angles)
angle = get_smallest_skew(img_resized, sigma_des, angles, map=map, logger=logger, plotter=plotter)
elif main_page:
angles = np.linspace(-12, 12, n_tot_angles)#np.array([0 , 45 , 90 , -45])
angles = np.array (list(np.linspace(-12, -7, int(n_tot_angles/4))) + list(np.linspace(-6, 6, n_tot_angles- 2* int(n_tot_angles/4))) + list(np.linspace(7, 12, int(n_tot_angles/4))))#np.linspace(-12, 12, n_tot_angles)#np.array([0 , 45 , 90 , -45])
angle = get_smallest_skew(img_resized, sigma_des, angles, map=map, logger=logger, plotter=plotter)
early_slope_edge=11