| import easyocr | |
| import numpy as np | |
| from PIL import Image | |
| def apply_ocr_remaining_area(remaining_area_image): | |
| reader = easyocr.Reader(['vi']) | |
| text = reader.readtext(np.array(remaining_area_image), detail=0) | |
| return ' '.join(text) | |
| def apply_ocr(cell_coordinates, cropped_table): | |
| reader = easyocr.Reader(['vi']) | |
| extracted_data = [] | |
| for row in cell_coordinates: | |
| row_data = [] | |
| for cell in row: | |
| x_min, y_min, x_max, y_max = cell['bbox'] | |
| cell_image = cropped_table.crop((x_min, y_min, x_max, y_max)) | |
| text = reader.readtext(np.array(cell_image), detail=0) | |
| row_data.append(' '.join(text)) | |
| extracted_data.append(row_data) | |
| return extracted_data | |