Update app_pages/ocr_comparator.py

app_pages/ocr_comparator.py

@@ -620,8 +620,8 @@ def app():
         # Recognize with Tesseract
         with st.spinner('Tesseract Text recognition in progress ...'):
             out_df_results_tesseract, status_tesseract = \
-            #tesserocr_recog(in_image_cv, in_list_dict_params[3], len(list_cropped_images))
             tesserocr_recog(in_image_cv, in_list_dict_params[2], len(list_cropped_images))
+            #tesserocr_recog(in_image_cv, in_list_dict_params[3], len(list_cropped_images))
         ##
 
         # Create results data frame
@@ -757,44 +757,44 @@ def app():
 #
     #    return out_list_text_mmocr, out_list_confidence_mmocr, out_status
 #
-    ####
-    #@st.experimental_memo(suppress_st_warning=True, show_spinner=False)
-    #def tesserocr_recog(in_img, in_params, in_nb_images):
-    #    """Recognition with Tesseract
-#
-    #    Args:
-    #        in_image_cv (matrix) : original image
-    #        in_params (dict)     : parameters for recognition
-    #        in_nb_images         : nb cropped images (used for progress bar)
-#
-    #    Returns:
-    #        Pandas data frame : recognition results
-    #        string/Exception  : recognition status
-    #    """
-    #    ## ------- Tesseract Text recognition
-    #    step = 3*in_nb_images # fourth recognition process
-    #    nb_steps = 4 * in_nb_images
-    #    progress_bar = st.progress(step/nb_steps)
-#
-    #    try:
-    #        out_df_result = pytesseract.image_to_data(in_img, **in_params,output_type=Output.DATAFRAME)
-#
-    #        out_df_result['box'] = out_df_result.apply(lambda d: [[d['left'], d['top']], \
-    #                                                    [d['left'] + d['width'], d['top']], \
-    #                                                    [d['left']+d['width'], d['top']+d['height']], \
-    #                                                    [d['left'], d['top'] + d['height']], \
-    #                                                    ], axis=1)
-    #        out_df_result['cropped'] = out_df_result['box'].apply(lambda b: cropped_1box(b, in_img))
-    #        out_df_result = out_df_result[(out_df_result.word_num > 0) & (out_df_result.text != ' ')] \
-    #                            .reset_index(drop=True)
-    #        out_status = 'OK'
-    #    except Exception as e:
-    #        out_df_result = pd.DataFrame([])
-    #        out_status = e
-#
-    #    progress_bar.progress(1.)
-#
-    #    return out_df_result, out_status
+    ###
+    @st.experimental_memo(suppress_st_warning=True, show_spinner=False)
+    def tesserocr_recog(in_img, in_params, in_nb_images):
+        """Recognition with Tesseract
+
+        Args:
+            in_image_cv (matrix) : original image
+            in_params (dict)     : parameters for recognition
+            in_nb_images         : nb cropped images (used for progress bar)
+
+        Returns:
+            Pandas data frame : recognition results
+            string/Exception  : recognition status
+        """
+        ## ------- Tesseract Text recognition
+        step = 3*in_nb_images # fourth recognition process
+        nb_steps = 4 * in_nb_images
+        progress_bar = st.progress(step/nb_steps)
+
+        try:
+            out_df_result = pytesseract.image_to_data(in_img, **in_params,output_type=Output.DATAFRAME)
+
+            out_df_result['box'] = out_df_result.apply(lambda d: [[d['left'], d['top']], \
+                                                        [d['left'] + d['width'], d['top']], \
+                                                        [d['left']+d['width'], d['top']+d['height']], \
+                                                        [d['left'], d['top'] + d['height']], \
+                                                        ], axis=1)
+            out_df_result['cropped'] = out_df_result['box'].apply(lambda b: cropped_1box(b, in_img))
+            out_df_result = out_df_result[(out_df_result.word_num > 0) & (out_df_result.text != ' ')] \
+                                .reset_index(drop=True)
+            out_status = 'OK'
+        except Exception as e:
+            out_df_result = pd.DataFrame([])
+            out_status = e
+
+        progress_bar.progress(1.)
+
+        return out_df_result, out_status
 
     ###
     def draw_reco_images(in_image, in_boxes_coordinates, in_list_texts, in_list_confid, \