HTR/hcr.py

from transformers import TrOCRProcessor, VisionEncoderDecoderModel
from PIL import Image
import cv2
import os
import torch
import sys
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from utils import notification_queue, log_print

# Global variables for model and processor
processor = None
model = None

def initialize_model():
    """Initialize the TrOCR model and processor"""
    global processor, model
    MODEL_NAME = "microsoft/trocr-large-handwritten"
    try:
        log_print("Initializing TrOCR model...")
        processor = TrOCRProcessor.from_pretrained(MODEL_NAME)
        model = VisionEncoderDecoderModel.from_pretrained(MODEL_NAME)
        if torch.cuda.is_available():
            model = model.to('cuda')
            log_print("Model moved to CUDA")
        log_print("TrOCR model initialized successfully")
    except Exception as e:
        error_msg = str(e)
        log_print(f"Error initializing TrOCR model: {error_msg}", "ERROR")
        raise

def text(image_cv):
    try:
        # Initialize model if not already initialized
        if processor is None or model is None:
            initialize_model()
            
        if not isinstance(image_cv, list):
            image_cv = [image_cv]
            
        t = ""
        total_images = len(image_cv)
        notification_queue.put({
            "type": "info",
            "message": f"Processing {total_images} image(s)..."
        })
        for i in image_cv:
            img_rgb = cv2.cvtColor(i, cv2.COLOR_BGR2RGB)
            image = Image.fromarray(img_rgb)

            pixel_values = processor(image, return_tensors="pt").pixel_values
            generated_ids = model.generate(pixel_values)

            generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
            t = t + generated_text.replace(" ", "") + " "
        return t
        
    except Exception as e:
        error_msg = str(e)
        notification_queue.put({
            "type": "error",
            "message": f"Error in text function: {error_msg}"
        })
        return ""