aiinnovators4/MobileNetV2

1from flask import Flask, request, render_template2from PIL import Image3import numpy as np4import tensorflow as tf5from tensorflow.keras.models import load_model6from tensorflow.keras.preprocessing import image as img_prep7import os8import tempfile9app = Flask(__name__)10 11# Load your trained model12model = load_model('animal_recognition_model.h5')13 14# Define class labels15classes = [16    'ArmaDillo', 'Bear', 'Birds', 'Cow', 'Crocodile', 'Deer', 'Elephant',17    'Goat', 'Horse', 'Jaguar', 'Monkey', 'Rabbit', 'Skunk', 'Tiger', 'Wild Boar'18]19 20# Image preprocessing function21def preprocess_image(image_path):22    try:23        img = Image.open(image_path)24        if img.mode != 'RGB':25            img = img.convert('RGB')26        img = img.resize((224, 224))  # Adjust to model input size27        img = img_prep.img_to_array(img)28        img = np.expand_dims(img, axis=0)29        img = img / 255.030        return img31    except Exception as e:32        print("Error processing image:", e)33        return None34 35# Prediction function36def predict(image_path):37    img = preprocess_image(image_path)38    if img is None:39        return "Image processing failed", "null"40 41    prediction = model.predict(img)42    confidence = np.max(prediction)43    if confidence > 0.85:44        predicted_class = classes[np.argmax(prediction)]45        return predicted_class, int(confidence * 100)46    else:47        return "No animal predicted", "null"48 49@app.route('/')50def home():51    return render_template('index.html')  # Ensure this HTML file exists52 53@app.route('/predict', methods=['POST'])54def predict_route():55    if 'file' not in request.files:56        return "No file part in the request"57    58    file = request.files['file']59    if file.filename == '':60        return "No selected file"61 62    with tempfile.NamedTemporaryFile(delete=False, suffix=".jpg") as tmp:63        image_path = tmp.name64        file.save(image_path)65 66    67    predicted_class, confidence = predict(image_path)68 69    # Clean up temporary file70    if os.path.exists(image_path):71        os.remove(image_path)72    73    return f"""74    Predicted Class: {predicted_class}75    Confidence: {confidence}%76    """77 78if __name__ == '__main__':79    app.run(host="0.0.0.0", port=7860, debug=True)80