TIF_NGANJUK_E41212268/app.py

from flask import Flask, render_template, request, redirect, url_for, jsonify
import os
import base64
import tensorflow as tf
from tensorflow import keras
from keras import preprocessing
import numpy as np
import cv2
from werkzeug.utils import secure_filename
import time

app = Flask(__name__)
app.config["UPLOAD_FOLDER"] = "static/uploads/"

# folder static/uploads
if not os.path.exists(app.config["UPLOAD_FOLDER"]):
    os.makedirs(app.config["UPLOAD_FOLDER"])

# Load model Hard Cascade Classifier
tomato_cascade = cv2.CascadeClassifier("model/tomato_classifier.xml")

# Load kedua model klasifikasi
#model_ripe = keras.models.load_model("model/CNN MobileNetV2-epochs10new-100.0.h5")
#model_fresh = keras.models.load_model("model/CNN MobileNetV2-epochs10kesegaran-99.66.h5")

modl_newgen = keras.models.load_model("model/CNN MobileNetV2-epochs5newgen-99.41.h5")

# Fungsi deteksi tomat
def is_tomato(img_path):
    """Deteksi apakah gambar mengandung tomat atau tidak"""
    img = cv2.imread(img_path)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    tomatoes = tomato_cascade.detectMultiScale(gray, scaleFactor=1.1, minNeighbors=5, minSize=(30, 30))
    return len(tomatoes) > 0  

# Fungsi prediksi gambar
def predict_image(img_path, model, class_names):
    img = preprocessing.image.load_img(img_path, target_size=(224, 224))
    img_array = preprocessing.image.img_to_array(img)
    img_array = np.expand_dims(img_array, axis=0)
    img_array /= 255.0
    predictions = model.predict(img_array)
    result = np.argmax(predictions, axis=1)
    confidence = int(np.max(predictions) * 100)
    return class_names[result[0]], confidence

@app.route("/")
def index():
    return render_template("index.html")

@app.route("/upload", methods=["GET", "POST"])
def upload():
    if request.method == "POST":
        # Proses file unggahan
        if 'file' not in request.files:
            return render_template("upload.html", error="Tidak ada file yang dipilih.")  
        
        file = request.files['file']
        if file.filename == '':
            return render_template("upload.html", error="Nama file kosong. Silakan pilih file lain.") 

        if file:
            filename = secure_filename(file.filename)
            filepath = os.path.join(app.config['UPLOAD_FOLDER'], filename)
            file.save(filepath)

            # Validasi gambar menggunakan Hard Cascade
            if not is_tomato(filepath):
                return render_template("upload.html", error="Gambar yang diunggah bukan tomat. Harap unggah gambar tomat.")

            # Simpan path gambar di sesi sementara untuk ditampilkan di halaman hasil
            return redirect(url_for("predict", img_path=filename))
    
    return render_template("upload.html")

@app.route("/predict", methods=["GET", "POST"])
def predict():
    if request.method == "POST":
        # Tangani data gambar dari kamera
        data = request.get_json()
        if data and "image" in data:
            # Decode Base64 menjadi file gambar
            image_data = data["image"].split(",")[1]  
            image_bytes = base64.b64decode(image_data)
            filename = f"camera_{int(time.time())}.png"
            filepath = os.path.join(app.config["UPLOAD_FOLDER"], filename)

            # Simpan gambar ke folder uploads
            with open(filepath, "wb") as f:
                f.write(image_bytes)

            # Validasi gambar menggunakan Hard Cascade
            if not is_tomato(filepath):
                return jsonify({"success": False, "error": "Gambar yang diambil bukan tomat. Harap ambil gambar tomat."}), 400

            # Lakukan prediksi setelah validasi
            #class_names_ripe = ['Belum Matang', 'Matang']
            #class_names_fresh = ['Busuk', 'Segar']
            
            class_names_tomat = ['Matang Busuk', 'Matang Segar', 'Mentah Busuk', 'Mentah Segar']

            # Prediksi kematangan
            #predicted_class_ripe, confidence_ripe = predict_image(filepath, model_ripe, class_names_ripe)

            # Prediksi kesegaran
            #predicted_class_fresh, confidence_fresh = predict_image(filepath, model_fresh, class_names_fresh)

            predicted_class_tomat, confidence_tomat = predict_image(filepath, modl_newgen, class_names_tomat)
            
            # Gabungkan hasil prediksi
            #predicted_class = f"{predicted_class_tomat}"
            #average_confidence = round((confidence_ripe + confidence_fresh) / 2, 2)

            # Kirim hasil prediksi sebagai respons JSON
            return jsonify({
                "success": True,
                "filename": filename,
                "predicted_class": predicted_class_tomat,
                "confidence": confidence_tomat
            })

        return jsonify({"success": False, "error": "No valid image data received"}), 400

    # Tangani metode GET untuk menampilkan halaman prediksi
    img_path = request.args.get("img_path")
    if not img_path:
        return redirect(url_for("upload"))

    filepath = os.path.join(app.config["UPLOAD_FOLDER"], img_path)
    if not os.path.exists(filepath):
        return redirect(url_for("upload"))

    #class_names_ripe = ['Belum Matang', 'Matang']
    #class_names_fresh = ['Busuk', 'Segar']
    
    class_names_tomat = ['Matang Busuk', 'Matang Segar', 'Mentah Busuk', 'Mentah Segar']

    # Prediksi kematangan
    #predicted_class_ripe, confidence_ripe = predict_image(filepath, model_ripe, class_names_ripe)

    # Prediksi kesegaran
    #predicted_class_fresh, confidence_fresh = predict_image(filepath, model_fresh, class_names_fresh)

    #predicted_class = f"{predicted_class_ripe} & {predicted_class_fresh}"
    #average_confidence = round((confidence_ripe + confidence_fresh) / 2, 2)

    predicted_class_tomat, confidence_tomat = predict_image(filepath, modl_newgen, class_names_tomat)
    
    local_time = time.localtime(time.time())
    data_date = time.strftime("%d-%m-%Y", local_time)
    data_time = time.strftime("%H:%M:%S", local_time)

    return render_template(
        "predict.html",
        img_path=img_path,
        predicted_class=predicted_class_tomat,
        confidence=confidence_tomat,
        data_date=data_date,
        data_time=data_time
    )

@app.route("/more")
def more():
    return render_template("more.html")

@app.route("/faq")
def faq():
    return render_template("faq.html")

@app.route("/about")
def about():
    return render_template("about.html")

if __name__ == "__main__":
    app.run(host="0.0.0.0", port=5000, debug=True)