test omitting PPOCR
Browse files- app_pages/ocr_comparator.py +23 -17
- requirements.txt +2 -2
app_pages/ocr_comparator.py
CHANGED
|
@@ -246,10 +246,12 @@ def app():
|
|
| 246 |
with st.spinner("EasyOCR reader initialization in progress ..."):
|
| 247 |
reader_easyocr = init_easyocr([in_list_params[0][0]])
|
| 248 |
|
|
|
|
| 249 |
# - PPOCR
|
| 250 |
# Paddleocr
|
| 251 |
with st.spinner("PPOCR reader initialization in progress ..."):
|
| 252 |
reader_ppocr = init_ppocr(in_list_params[1])
|
|
|
|
| 253 |
|
| 254 |
# - MMOCR
|
| 255 |
with st.spinner("MMOCR reader initialization in progress ..."):
|
|
@@ -470,7 +472,7 @@ def app():
|
|
| 470 |
else:
|
| 471 |
easyocr_boxes_coordinates = easyocr_status
|
| 472 |
##
|
| 473 |
-
|
| 474 |
## ------- PPOCR Text detection
|
| 475 |
with st.spinner('PPOCR Text detection in progress ...'):
|
| 476 |
ppocr_boxes_coordinates, ppocr_status = ppocr_detect(_in_list_readers[1], in_image_path)
|
|
@@ -481,6 +483,7 @@ def app():
|
|
| 481 |
else:
|
| 482 |
ppocr_image_detect = ppocr_status
|
| 483 |
##
|
|
|
|
| 484 |
|
| 485 |
## ------- MMOCR Text detection
|
| 486 |
with st.spinner('MMOCR Text detection in progress ...'):
|
|
@@ -589,8 +592,8 @@ def app():
|
|
| 589 |
|
| 590 |
list_text_easyocr = []
|
| 591 |
list_confidence_easyocr = []
|
| 592 |
-
list_text_ppocr = []
|
| 593 |
-
list_confidence_ppocr = []
|
| 594 |
list_text_mmocr = []
|
| 595 |
list_confidence_mmocr = []
|
| 596 |
|
|
@@ -602,12 +605,13 @@ def app():
|
|
| 602 |
list_text_easyocr, list_confidence_easyocr, status_easyocr = \
|
| 603 |
easyocr_recog(list_cropped_images, in_list_readers[0], in_list_dict_params[0])
|
| 604 |
##
|
| 605 |
-
|
| 606 |
# Recognize with PPOCR
|
| 607 |
with st.spinner('PPOCR Text recognition in progress ...'):
|
| 608 |
list_text_ppocr, list_confidence_ppocr, status_ppocr = \
|
| 609 |
ppocr_recog(list_cropped_images, in_list_dict_params[1])
|
| 610 |
##
|
|
|
|
| 611 |
|
| 612 |
# Recognize with MMOCR
|
| 613 |
with st.spinner('MMOCR Text recognition in progress ...'):
|
|
@@ -625,14 +629,15 @@ def app():
|
|
| 625 |
out_df_results = pd.DataFrame({'cropped_image': list_cropped_images,
|
| 626 |
'text_easyocr': list_text_easyocr,
|
| 627 |
'confidence_easyocr': list_confidence_easyocr,
|
| 628 |
-
'text_ppocr': list_text_ppocr,
|
| 629 |
-
'confidence_ppocr': list_confidence_ppocr,
|
| 630 |
'text_mmocr': list_text_mmocr,
|
| 631 |
'confidence_mmocr': list_confidence_mmocr
|
| 632 |
}
|
| 633 |
)
|
| 634 |
|
| 635 |
-
out_list_reco_status = [status_easyocr, status_ppocr, status_mmocr, status_tesseract]
|
|
|
|
| 636 |
|
| 637 |
return out_df_results, out_df_results_tesseract, out_list_reco_status
|
| 638 |
|
|
@@ -674,21 +679,21 @@ def app():
|
|
| 674 |
progress_bar.empty()
|
| 675 |
|
| 676 |
return out_list_text_easyocr, out_list_confidence_easyocr, out_status
|
| 677 |
-
|
| 678 |
###
|
| 679 |
#@st.experimental_memo(suppress_st_warning=True, show_spinner=False)
|
| 680 |
@st.cache_data
|
| 681 |
def ppocr_recog(in_list_images, in_params):
|
| 682 |
-
"""Recognition with PPOCR
|
| 683 |
|
| 684 |
-
Args:
|
| 685 |
-
|
| 686 |
-
|
| 687 |
|
| 688 |
-
Returns:
|
| 689 |
-
|
| 690 |
-
|
| 691 |
-
|
| 692 |
"""
|
| 693 |
## ------- PPOCR Text recognition
|
| 694 |
out_list_text_ppocr = []
|
|
@@ -714,6 +719,7 @@ def app():
|
|
| 714 |
progress_bar.empty()
|
| 715 |
|
| 716 |
return out_list_text_ppocr, out_list_confidence_ppocr, out_status
|
|
|
|
| 717 |
|
| 718 |
###
|
| 719 |
#@st.experimental_memo(suppress_st_warning=True, show_spinner=False)
|
|
@@ -951,7 +957,7 @@ def app():
|
|
| 951 |
process_detect.clear()
|
| 952 |
get_cropped.clear()
|
| 953 |
easyocr_recog.clear()
|
| 954 |
-
ppocr_recog.clear()
|
| 955 |
mmocr_recog.clear()
|
| 956 |
tesserocr_recog.clear()
|
| 957 |
|
|
|
|
| 246 |
with st.spinner("EasyOCR reader initialization in progress ..."):
|
| 247 |
reader_easyocr = init_easyocr([in_list_params[0][0]])
|
| 248 |
|
| 249 |
+
"""
|
| 250 |
# - PPOCR
|
| 251 |
# Paddleocr
|
| 252 |
with st.spinner("PPOCR reader initialization in progress ..."):
|
| 253 |
reader_ppocr = init_ppocr(in_list_params[1])
|
| 254 |
+
"""
|
| 255 |
|
| 256 |
# - MMOCR
|
| 257 |
with st.spinner("MMOCR reader initialization in progress ..."):
|
|
|
|
| 472 |
else:
|
| 473 |
easyocr_boxes_coordinates = easyocr_status
|
| 474 |
##
|
| 475 |
+
"""
|
| 476 |
## ------- PPOCR Text detection
|
| 477 |
with st.spinner('PPOCR Text detection in progress ...'):
|
| 478 |
ppocr_boxes_coordinates, ppocr_status = ppocr_detect(_in_list_readers[1], in_image_path)
|
|
|
|
| 483 |
else:
|
| 484 |
ppocr_image_detect = ppocr_status
|
| 485 |
##
|
| 486 |
+
"""
|
| 487 |
|
| 488 |
## ------- MMOCR Text detection
|
| 489 |
with st.spinner('MMOCR Text detection in progress ...'):
|
|
|
|
| 592 |
|
| 593 |
list_text_easyocr = []
|
| 594 |
list_confidence_easyocr = []
|
| 595 |
+
#list_text_ppocr = []
|
| 596 |
+
#list_confidence_ppocr = []
|
| 597 |
list_text_mmocr = []
|
| 598 |
list_confidence_mmocr = []
|
| 599 |
|
|
|
|
| 605 |
list_text_easyocr, list_confidence_easyocr, status_easyocr = \
|
| 606 |
easyocr_recog(list_cropped_images, in_list_readers[0], in_list_dict_params[0])
|
| 607 |
##
|
| 608 |
+
"""
|
| 609 |
# Recognize with PPOCR
|
| 610 |
with st.spinner('PPOCR Text recognition in progress ...'):
|
| 611 |
list_text_ppocr, list_confidence_ppocr, status_ppocr = \
|
| 612 |
ppocr_recog(list_cropped_images, in_list_dict_params[1])
|
| 613 |
##
|
| 614 |
+
"""
|
| 615 |
|
| 616 |
# Recognize with MMOCR
|
| 617 |
with st.spinner('MMOCR Text recognition in progress ...'):
|
|
|
|
| 629 |
out_df_results = pd.DataFrame({'cropped_image': list_cropped_images,
|
| 630 |
'text_easyocr': list_text_easyocr,
|
| 631 |
'confidence_easyocr': list_confidence_easyocr,
|
| 632 |
+
#'text_ppocr': list_text_ppocr,
|
| 633 |
+
#'confidence_ppocr': list_confidence_ppocr,
|
| 634 |
'text_mmocr': list_text_mmocr,
|
| 635 |
'confidence_mmocr': list_confidence_mmocr
|
| 636 |
}
|
| 637 |
)
|
| 638 |
|
| 639 |
+
#out_list_reco_status = [status_easyocr, status_ppocr, status_mmocr, status_tesseract]
|
| 640 |
+
out_list_reco_status = [status_easyocr, status_mmocr, status_tesseract]
|
| 641 |
|
| 642 |
return out_df_results, out_df_results_tesseract, out_list_reco_status
|
| 643 |
|
|
|
|
| 679 |
progress_bar.empty()
|
| 680 |
|
| 681 |
return out_list_text_easyocr, out_list_confidence_easyocr, out_status
|
| 682 |
+
"""
|
| 683 |
###
|
| 684 |
#@st.experimental_memo(suppress_st_warning=True, show_spinner=False)
|
| 685 |
@st.cache_data
|
| 686 |
def ppocr_recog(in_list_images, in_params):
|
| 687 |
+
"""#Recognition with PPOCR
|
| 688 |
|
| 689 |
+
#Args:
|
| 690 |
+
# in_list_images (list) : list of cropped images
|
| 691 |
+
# in_params (dict) : parameters for recognition
|
| 692 |
|
| 693 |
+
#Returns:
|
| 694 |
+
# list : list of recognized text
|
| 695 |
+
# list : list of recognition confidence
|
| 696 |
+
# string/Exception : recognition status
|
| 697 |
"""
|
| 698 |
## ------- PPOCR Text recognition
|
| 699 |
out_list_text_ppocr = []
|
|
|
|
| 719 |
progress_bar.empty()
|
| 720 |
|
| 721 |
return out_list_text_ppocr, out_list_confidence_ppocr, out_status
|
| 722 |
+
"""
|
| 723 |
|
| 724 |
###
|
| 725 |
#@st.experimental_memo(suppress_st_warning=True, show_spinner=False)
|
|
|
|
| 957 |
process_detect.clear()
|
| 958 |
get_cropped.clear()
|
| 959 |
easyocr_recog.clear()
|
| 960 |
+
#ppocr_recog.clear()
|
| 961 |
mmocr_recog.clear()
|
| 962 |
tesserocr_recog.clear()
|
| 963 |
|
requirements.txt
CHANGED
|
@@ -9,11 +9,11 @@ Pillow
|
|
| 9 |
mmcv-full --no-binary mmcv-full
|
| 10 |
mmdet==2.28.2
|
| 11 |
mmocr==0.6.3
|
| 12 |
-
paddlepaddle==2.4.0rc0
|
| 13 |
numpy==1.23.4
|
| 14 |
mycolorpy==1.5.1
|
| 15 |
plotly==5.10.0
|
| 16 |
plotly-express==0.4.1
|
| 17 |
pytesseract==0.3.10
|
| 18 |
streamlit_option_menu
|
| 19 |
-
paddleocr
|
|
|
|
| 9 |
mmcv-full --no-binary mmcv-full
|
| 10 |
mmdet==2.28.2
|
| 11 |
mmocr==0.6.3
|
| 12 |
+
#paddlepaddle==2.4.0rc0
|
| 13 |
numpy==1.23.4
|
| 14 |
mycolorpy==1.5.1
|
| 15 |
plotly==5.10.0
|
| 16 |
plotly-express==0.4.1
|
| 17 |
pytesseract==0.3.10
|
| 18 |
streamlit_option_menu
|
| 19 |
+
#paddleocr
|