import easyocr
import cv2
# Initialize EasyOCR
reader = easyocr.Reader(['en'])
# Load the image
image = cv2.imread('book.jpg')
# Run OCR directly
results = reader.readtext(image)
# Display results
for (bbox, text, prob) in results:
    print(f"Detected Text: {text} (Probability: {prob})")

THE 0 R |G |NAL B E STSELLE R THE SECRET HISTORY DONNA TARTT Haunting, compelling and brilliant The Times

from ultralytics import YOLO

model = YOLO("yolov10n.pt")
# Train the model
model.train(data="datasets/data.yaml", epochs=50, imgsz=640)

import cv2
# Open video file
video_path = 'books.mov'
cap = cv2.VideoCapture(video_path)
# Load YOLO model
model = YOLO('model.pt')
# Function for object detection and drawing bounding boxes
def predict_and_detect(model, frame, conf=0.5):
    results = model.predict(frame, conf=conf)
    for result in results:
        for box in result.boxes:
            # Draw bounding box
            x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())
            cv2.rectangle(frame, (x1, y1), (x2, y2), (255, 0, 0), 2)
    return frame, results
# Process video frames
while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break
    # Run object detection
    processed_frame, results = predict_and_detect(model, frame)
    # Show video with bounding boxes
    cv2.imshow('YOLO + OCR Detection', processed_frame)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break
# Release video
cap.release()
cv2.destroyAllWindows()

import easyocr
# Initialize EasyOCR
reader = easyocr.Reader(['en'])
# Function to crop frames and perform OCR
def run_ocr_on_boxes(frame, boxes):
    ocr_results = []
    for box in boxes:
        x1, y1, x2, y2 = map(int, box.xyxy[0].tolist())
        cropped_frame = frame[y1:y2, x1:x2]
        ocr_result = reader.readtext(cropped_frame)
        ocr_results.append(ocr_result)
    return ocr_results
# Perform OCR on detected bounding boxes
for result in results:
    ocr_results = run_ocr_on_boxes(frame, result.boxes)
    # Extract and display the text from OCR results
    extracted_text = [detection[1] for ocr in ocr_results for detection in ocr]
    print(f"Extracted Text: {', '.join(extracted_text)}")

'THE, SECRET, HISTORY, DONNA, TARTT'

import ollama
# Construct a prompt to clean up the OCR output
prompt = f"""
- Below is a text extracted from an OCR. The text contains mentions of famous books and their corresponding authors.
- Some words may be slightly misspelled or out of order.
- Your task is to identify the book titles and corresponding authors from the text.
- Output the text in the format: '<Name of the book> : <Name of the author>'.
- Do not generate any other text except the book title and the author.
TEXT:
{output_text}
"""
# Use Ollama to clean and structure the OCR output
response = ollama.chat(
    model="llama3",
    messages=[{"role": "user", "content": prompt}]
)
# Extract cleaned text
cleaned_text = response['message']['content'].strip()
print(cleaned_text)

The Secret History : Donna Tartt