import os from PIL import Image, ImageOps, ImageChops import io import fitz # PyMuPDF from docx import Document from rembg import remove import gradio as gr import os import os from io import BytesIO from docx import Document from PIL import Image # ایجاد دایرکتوری static os.makedirs("static", exist_ok=True) def trim_whitespace(image): gray_image = ImageOps.grayscale(image) inverted_image = ImageChops.invert(gray_image) bbox = inverted_image.getbbox() trimmed_image = image.crop(bbox) return trimmed_image def convert_pdf_to_images(pdf_path, zoom=2): pdf_document = fitz.open(pdf_path) images = [] for page_num in range(len(pdf_document)): page = pdf_document.load_page(page_num) matrix = fitz.Matrix(zoom, zoom) pix = page.get_pixmap(matrix=matrix) image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples) trimmed_image = trim_whitespace(image) images.append(trimmed_image) return images def convert_docx_to_jpeg(docx_bytes): """ Convert each image in a DOCX file to a separate JPEG image and return them as a list. Args: - docx_bytes: The binary content of the DOCX file. Returns: - A list of PIL Image objects in JPEG format. """ document = Document(BytesIO(docx_bytes)) images = [] for rel in document.part.rels.values(): if "image" in rel.target_ref: image_stream = rel.target_part.blob image = Image.open(BytesIO(image_stream)) jpeg_image = BytesIO() image.convert('RGB').save(jpeg_image, format="JPEG") jpeg_image.seek(0) images.append(Image.open(jpeg_image)) return images # Example usage: # with open("example.docx", "rb") as f: # docx_bytes = f.read() # images = convert_docx_to_jpeg(docx_bytes) # for img in images: # img.show() def remove_background_from_image(image): return remove(image) def process_file(input_file): file_extension = os.path.splitext(input_file.name)[1].lower() if file_extension in ['.png', '.jpeg', '.jpg', '.bmp', '.gif']: image = Image.open(input_file) image = image.convert('RGB') output_image = remove_background_from_image(image) return output_image elif file_extension == '.pdf': images = convert_pdf_to_images(input_file.name) return [remove_background_from_image(image) for image in images] elif file_extension in ['.docx', '.doc']: images = convert_docx_to_jpeg(input_file.name) return [remove_background_from_image(image) for image in images] else: return "File format not supported." def gradio_interface(input_file): return process_file(input_file) iface = gr.Interface( fn=gradio_interface, inputs=gr.File(label="Upload Word, PDF, or Image"), outputs=gr.Image(type="pil", label="Processed Image(s)"), title="Document to Image Converter with Background Removal" ) if __name__ == "__main__": iface.launch()