' img_url = img_tag[img_tag.index('src=')+5:img_tag.index('.jpg')+4] response = requests.get(img_url) image = cv2.imdecode(np.frombuffer(response.content, np.uint8), cv2.IMREAD_COLOR) # Convert the image to grayscale and apply Otsu's thresholding gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) threshold, binary_image = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU) # Remove noise using dilation and erosion kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3,3)) dilated_image = cv2.dilate(binary_image, kernel, iterations=1) eroded_image = cv2.erode(dilated_image, kernel, iterations=1) # Extract text from the image using pytesseract text = pytesseract.image_to_string(eroded_image) # Replace the img tag with the extracted text new_post_content = img_tag.replace(img_tag, text) # Use the Blogger API to update the post with the new content creds = Credentials.from_authorized_user_info(info=None) service = build('blogger', 'v3', credentials=creds) request = service.posts().patch(blogId=BLOG_ID, postId='POST_ID', body={'content': new_post_content}) response = request.execute()

Search This Blog

advertisement

Posts