app.py

from flask import Flask, request, jsonify, send_file, render_template_string
from flask_cors import CORS
from ultralytics import YOLO
import cv2
import os
import uuid
import json
from werkzeug.utils import secure_filename

# ---------------------------
# Initialize Flask app
# ---------------------------
app = Flask(__name__)
CORS(app)

# ---------------------------
# Load YOLO model (ONLY ONCE)
# ---------------------------
model = YOLO("best.pt")

# ---------------------------
# Create upload/result folders
# ---------------------------
UPLOAD_FOLDER = "uploads"
RESULTS_FOLDER = "results"

os.makedirs(UPLOAD_FOLDER, exist_ok=True)
os.makedirs(RESULTS_FOLDER, exist_ok=True)

# ---------------------------
# Load fertilizer database
# ---------------------------
with open("fertilizer_data.json", "r") as f:
    fertilizer_db = json.load(f)

# ---------------------------
# Health Check Route (Important for Render)
# ---------------------------
@app.route("/health")
def health():
    return {"status": "Backend Running Successfully ✅"}

# ---------------------------
# Home Page Route
# ---------------------------
@app.route("/")
def home():
    return render_template_string("""
        <h2>🚀 YOLO Flask API with Fertilizer Recommendation</h2>
        <p>Upload an image to detect weeds and get fertilizer suggestions.</p>

        <form action="/predict" method="post" enctype="multipart/form-data">
            <input type="file" name="file">
            <input type="submit" value="Upload & Detect">
        </form>
    """)

# ---------------------------
# Prediction API Route
# ---------------------------
@app.route("/predict", methods=["POST"])
def predict():
    try:
        # ✅ 1. Check uploaded file
        if "file" in request.files:
            file = request.files["file"]
        elif "image" in request.files:
            file = request.files["image"]
        else:
            return jsonify({"error": "No image uploaded"}), 400

        if file.filename == "":
            return jsonify({"error": "No image selected"}), 400

        # ✅ 2. Save uploaded image
        filename = secure_filename(file.filename)

        if not filename:
            filename = str(uuid.uuid4()) + ".jpg"

        filepath = os.path.join(UPLOAD_FOLDER, filename)
        file.save(filepath)

        # ✅ 3. Run YOLO Prediction
        results = model.predict(filepath)

        # ✅ 4. Read image for drawing
        img = cv2.imread(filepath)
        detections = []

        # ✅ 5. Loop over detected boxes
        for box in results[0].boxes:
            cls_id = int(box.cls[0])
            label = results[0].names[cls_id]
            conf = float(box.conf[0])

            # ✅ Fertilizer Info Fetch
            fert_info = fertilizer_db.get(label, {
                "fertilizer": "Not found",
                "quantity": "N/A",
                "frequency": "N/A"
            })

            # ✅ Add detection record
            detections.append({
                "label": label,
                "confidence": round(conf * 100, 2),
                "fertilizer": fert_info["fertilizer"],
                "quantity": fert_info["quantity"],
                "frequency": fert_info["frequency"]
            })

            # ✅ Draw bounding box
            x1, y1, x2, y2 = map(int, box.xyxy[0])
            cv2.rectangle(img, (x1, y1), (x2, y2), (0, 255, 0), 2)

            # ✅ Draw label text
            text = f"{label} {conf*100:.1f}%"
            cv2.putText(
                img,
                text,
                (x1, y1 - 10),
                cv2.FONT_HERSHEY_SIMPLEX,
                0.7,
                (255, 0, 0),
                2
            )

        # ✅ 6. Save Result Image
        result_filename = f"result_{filename}"
        result_path = os.path.join(RESULTS_FOLDER, result_filename)
        cv2.imwrite(result_path, img)

        # ✅ 7. Generate Image URLs
        base_url = request.host_url.rstrip("/")

        return jsonify({
            "detections": detections,
            "result_image_url": f"{base_url}/result/{result_filename}",
            "original_image_url": f"{base_url}/uploads/{filename}"
        })

    except Exception as e:
        print("Prediction Error:", str(e))
        return jsonify({"error": "Backend prediction failed", "details": str(e)}), 500


# ---------------------------
# Route for serving Result Image
# ---------------------------
@app.route("/result/<filename>")
def result_image(filename):
    return send_file(
        os.path.join(RESULTS_FOLDER, filename),
        mimetype="image/jpeg"
    )


# ---------------------------
# Route for serving Uploaded Image
# ---------------------------
@app.route("/uploads/<filename>")
def uploaded_image(filename):
    return send_file(
        os.path.join(UPLOAD_FOLDER, filename),
        mimetype="image/jpeg"
    )


# ---------------------------
# Main Run (Local only)
# ---------------------------
if __name__ == "__main__":
    app.run(host="0.0.0.0", port=5000)