Capstone Project: Build an AI-Powered Image Processing App with OpenCV

Chapter 5.5: Final Capstone Project – Build Your Own AI-Powered Image Processing App

It’s time to put everything together! 🎯 In this capstone project, you will build an AI-powered Image Processing App that combines different OpenCV features like filters, sketch effect, ANPR, and face/mask detection. This app will give you real-world, portfolio-ready experience in computer vision.

🚀 App Features

• 📷 Apply Instagram-like filters (grayscale, sepia, cartoon, negative)
• ✏️ Real-time sketch effect from webcam
• 🚗 Number Plate Detection (ANPR basics with OCR)
• 😷 Face Mask Detection with pre-trained model
• 🔄 Menu-driven interface for switching between modules

📌 Step 1: App Setup

import cv2
import numpy as np
import pytesseract
from tensorflow.keras.models import load_model

# Load models
face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_frontalface_default.xml")
mask_model = load_model("mask_detector_model.h5")

📌 Step 2: Menu System

def menu():
    print("Choose an option:")
    print("1. Apply Filters")
    print("2. Real-time Sketch")
    print("3. Number Plate Detection")
    print("4. Face Mask Detection")
    print("5. Exit")

📌 Step 3: Implement Features

Each menu option calls the relevant OpenCV function. Example: applying filters, detecting faces with masks, or detecting number plates.

def sketch_webcam():
    cap = cv2.VideoCapture(0)
    while True:
        ret, frame = cap.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        inv = cv2.bitwise_not(gray)
        blur = cv2.GaussianBlur(inv, (21, 21), 0)
        sketch = cv2.divide(gray, 255 - blur, scale=256)
        cv2.imshow("Sketch Effect", sketch)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    cap.release()
    cv2.destroyAllWindows()

def mask_detection():
    cap = cv2.VideoCapture(0)
    while True:
        ret, frame = cap.read()
        gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
        faces = face_cascade.detectMultiScale(gray, 1.1, 4)
        for (x,y,w,h) in faces:
            face = cv2.resize(frame[y:y+h, x:x+w], (128,128)) / 255.0
            face = np.expand_dims(face, axis=0)
            pred = mask_model.predict(face)
            label = "Mask" if pred[0][0] > 0.5 else "No Mask"
            color = (0,255,0) if label=="Mask" else (0,0,255)
            cv2.rectangle(frame,(x,y),(x+w,y+h),color,2)
            cv2.putText(frame,label,(x,y-10),cv2.FONT_HERSHEY_SIMPLEX,0.8,color,2)
        cv2.imshow("Mask Detection", frame)
        if cv2.waitKey(1) & 0xFF == ord('q'):
            break
    cap.release()
    cv2.destroyAllWindows()

💡 Applications

• Portfolio project for data science & computer vision interviews
• Educational tool to demonstrate AI & OpenCV capabilities
• Base for developing real-world apps (surveillance, AR filters, OCR)

✅ Congratulations! You’ve completed the Capstone Project for this course. By building an AI-powered Image Processing App, you have integrated filters, sketching, ANPR, and mask detection. This is a portfolio-ready project to showcase your OpenCV and AI skills.
👉 Next, we’ll wrap up with Quiz 5 to test your knowledge of advanced applications.