HTR/app.py

import cv2
import os
import tempfile
import logging
import numpy as np

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

from HTR.word import convert_image
from HTR.strike import struck_images
from HTR.hcr import text
from HTR.spell_and_gramer_check import spell_grammer

def preprocess_image(img):
    """Preprocess image to improve text detection"""
    try:
        # Convert to grayscale
        gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
        
        # Apply adaptive thresholding
        binary = cv2.adaptiveThreshold(
            gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, 
            cv2.THRESH_BINARY_INV, 11, 2
        )
        
        # Denoise
        denoised = cv2.fastNlMeansDenoising(binary)
        
        # Convert back to BGR
        return cv2.cvtColor(denoised, cv2.COLOR_GRAY2BGR)
    except Exception as e:
        logger.error(f"Error in preprocess_image: {str(e)}")
        return img

def extract_text_from_image(img_path):
    try:
        # Ensure the image exists
        if not os.path.exists(img_path):
            logger.error(f"Image file not found: {img_path}")
            return ""
            
        # Read the image
        img = cv2.imread(img_path)
        if img is None:
            logger.error(f"Failed to read image: {img_path}")
            return ""
            
        # Log image properties
        logger.info(f"Processing image: {img_path}")
        logger.info(f"Image shape: {img.shape}")
        logger.info(f"Image dtype: {img.dtype}")
        
        # Convert image to RGB if needed
        if len(img.shape) == 2:  # If grayscale
            img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR)
        elif img.shape[2] == 4:  # If RGBA
            img = cv2.cvtColor(img, cv2.COLOR_RGBA2BGR)
            
        # Ensure image is in correct format
        img = img.astype(np.uint8)
        
        # Preprocess image
        logger.info("Preprocessing image...")
        img = preprocess_image(img)
        
        # Process the image
        logger.info("Converting image...")
        imgs = convert_image(img)
        if not imgs:
            logger.error("No text regions detected in image")
            return ""
            
        logger.info(f"Found {len(imgs)} text regions")
        
        logger.info("Processing struck images...")
        images_path = struck_images(imgs)
        if not images_path:
            logger.error("No valid text regions after strike processing")
            return ""
            
        logger.info("Extracting text...")
        t = text(images_path)
        if not t:
            logger.error("No text extracted from image")
            return ""
            
        logger.info("Spell checking...")
        t = spell_grammer(t)
        
        logger.info(f"Extracted text: {t}")
        return t
        
    except Exception as e:
        logger.error(f"Error in extract_text_from_image: {str(e)}")
        return ""

# extract_text_from_image("ans_image/1.jpg")