import cv2 import numpy as np from torchvision import transforms def get_text_map_simple(image): if isinstance(image, str): image = cv2.imread(image) if image is None: return np.zeros((256, 256, 3)) # robustness gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3, 3)) gradient = cv2.morphologyEx(gray, cv2.MORPH_GRADIENT, kernel) _, binary = cv2.threshold(gradient, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU) text_map = cv2.cvtColor(binary, cv2.COLOR_GRAY2BGR) return text_map def get_transforms(img_size): return transforms.Compose([ transforms.Resize((img_size, img_size)), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]) ])