3.7 KiB

Raw Permalink Blame History

Tomat Classification API

API Flask untuk klasifikasi tingkat kematangan tomat menggunakan Random Forest dan Color Histogram RGB.

Fitur

Preprocessing: Resize gambar ke 256x256
Ekstraksi Fitur: Color Histogram RGB (8x8x8 bins)
Klasifikasi: Random Forest dengan 3 kelas
Format Output: JSON dengan probabilitas dan confidence score

Kelas Output

matang - Tomat sudah matang
mentah - Tomat masih mentah
setengah_matang - Tomat setengah matang

Endpoint

1. Health Check

GET /health

Response:

{
  "status": "healthy",
  "model_loaded": true,
  "service": "Tomat Classification API"
}

2. Prediksi

POST /predict
Content-Type: multipart/form-data

Request: Upload file gambar dengan key image

Response:

{
  "success": true,
  "prediction": {
    "class": "matang",
    "confidence": 0.85,
    "confidence_percentage": 85.0,
    "probabilities": {
      "matang": {"probability": 0.85, "percentage": 85.0},
      "mentah": {"probability": 0.10, "percentage": 10.0},
      "setengah_matang": {"probability": 0.05, "percentage": 5.0}
    }
  },
  "metadata": {
    "model_type": "RandomForest",
    "features_used": 24,
    "image_processed": "tomat.jpg"
  }
}

3. Informasi Model

GET /info

Response:

{
  "success": true,
  "model_info": {
    "type": "RandomForestClassifier",
    "classes": ["matang", "mentah", "setengah_matang"],
    "n_features": 24,
    "n_estimators": 100
  },
  "api_info": {
    "version": "1.0.0",
    "endpoints": {
      "health": "/health",
      "predict": "/predict (POST)",
      "info": "/info"
    },
    "supported_formats": ["PNG", "JPG", "JPEG"],
    "max_file_size": "16MB"
  }
}

Cara Menjalankan

1. Install Dependencies

pip install flask opencv-python numpy scikit-learn joblib

2. Training Model (Opsional)

python main.py

Ini akan membuat folder models/ dengan file model yang sudah trained.

3. Jalankan API

python app.py

Server akan berjalan di: http://127.0.0.1:5000

Cara Testing API

Menggunakan curl

# Health check
curl http://127.0.0.1:5000/health

# Prediksi gambar
curl -X POST -F "image=@path/to/gambar.jpg" http://127.0.0.1:5000/predict

# Info model
curl http://127.0.0.1:5000/info

Menggunakan Python

import requests

# Health check
response = requests.get('http://127.0.0.1:5000/health')
print(response.json())

# Prediksi gambar
with open('gambar_tomat.jpg', 'rb') as f:
    files = {'image': f}
    response = requests.post('http://127.0.0.1:5000/predict', files=files)
    print(response.json())

Struktur File

data_tomat/
  app.py                    # Flask API
  main.py                   # Training model
  models/                   # Folder model
    tomat_classifier.pkl    # Model Random Forest
    label_encoder.pkl       # Label encoder
    metadata.pkl            # Metadata model
  matang/                   # Folder gambar matang
  mentah/                   # Folder gambar mentah
  setengah_matang/          # Folder gambar setengah matang

Error Handling

API mengembalikan error response dengan format:

{
  "success": false,
  "error": "Error type",
  "message": "Detailed error message"
}

Common Errors

400: File tidak ada, format tidak didukung, processing gagal
413: File terlalu besar (>16MB)
500: Internal server error

Notes

API akan otomatis membuat model dummy jika file model belum ada
Untuk hasil prediksi yang akurat, jalankan python main.py terlebih dahulu
File temporary akan otomatis dihapus setelah processing
Gambar akan di-resize ke 256x256 sebelum ekstraksi fitur

Dependencies

Flask 2.0+
OpenCV 4.0+
NumPy 1.19+
Scikit-learn 1.0+
Joblib 1.0+

3.7 KiB Raw Permalink Blame History