app.py

import cv2
import streamlit as st
from fer import FER
from PIL import Image
import numpy as np


# Initialize emotion detector
emotion_detector = FER()

# Function to process image and detect emotions
def process_image(image):
    frame = np.array(image)
    frame = cv2.cvtColor(frame, cv2.COLOR_RGB2BGR)  # Convert from PIL to OpenCV format
    emotions = emotion_detector.detect_emotions(frame)
    blurred_frame = cv2.GaussianBlur(frame, (51, 51), 0)

    for face in emotions:
        (x, y, w, h) = face["box"]
        emotion, score = max(face["emotions"].items(), key=lambda item: item[1])

        overlay = frame.copy()
        alpha = 0.4
        cv2.rectangle(overlay, (x, y), (x + w, y + h), (0, 255, 0), 2)
        cv2.addWeighted(overlay, alpha, frame, 1 - alpha, 0, frame)

        blurred_frame[y:y + h, x:x + w] = frame[y:y + h, x:x + w]

        font = cv2.FONT_HERSHEY_SIMPLEX
        font_scale = 0.7
        font_thickness = 2
        text_color = (255, 255, 255)
        bg_color = (0, 0, 0)

        text = f"{emotion}: {score:.2f}"
        (text_width, text_height), _ = cv2.getTextSize(text, font, font_scale, font_thickness)
        text_x = x + 10
        text_y = y - 10 if y - 10 > text_height else y + h + text_height

        cv2.rectangle(blurred_frame, (text_x - 5, text_y - text_height - 5), (text_x + text_width + 5, text_y + 5), bg_color, -1)
        cv2.putText(blurred_frame, text, (text_x, text_y), font, font_scale, text_color, font_thickness)

    return cv2.cvtColor(blurred_frame, cv2.COLOR_BGR2RGB)  # Convert back to RGB for Streamlit

# Streamlit UI
st.title("Real-Time Emotion Recognition")
st.write("Use the camera or upload an image to detect emotions.")

# Camera Input
camera_image = st.camera_input("Take a picture")

# File Upload
uploaded_file = st.file_uploader("Or upload an image...", type=["jpg", "png", "jpeg"])

# Process image if uploaded or captured via camera
if camera_image or uploaded_file:
    image = Image.open(camera_image if camera_image else uploaded_file)
    st.image(image, caption="Captured Image", use_column_width=True)

    processed_image = process_image(image)
    st.image(processed_image, caption="Processed Image with Emotions", use_column_width=True)