def Extract_text_from_image(shape, Reader, Slide_number, image_number):
"""Извлечь текст из изображений в слайдах PPT"" "
image_bytes = shape.image.blob
image_stream = io.BytesIO(image_bytes)
img = Image.open(image_stream)
Код: Выделить всё
image_path = os.path.join(EXTRACTED_IMG_DIR, f"slide_{slide_number}_image_{image_number}.png")
img.save(image_path)
img_np = np.array(img.convert('L'))
result = reader.readtext(img_np, detail=1)
rows = []
current_row = []
previous_y = None
tolerance = 15
for entry in result:
bbox, text, confidence = entry
x_min, y_min = bbox[0]
if previous_y is None or abs(y_min - previous_y)
Подробнее здесь: [url]https://stackoverflow.com/questions/79184811/can-anyone-tell-how-can-i-convert-the-images-into-png-format[/url]