README.md

Palm Oil Ripeness Agent (n8n + YOLOv8)

This project uses a custom-trained YOLOv8 model to detect the ripeness of Palm Oil Fresh Fruit Bunches (FFB). It features a local Python FastAPI server that integrates into an agentic n8n workflow, storing results and embeddings in MongoDB Atlas.

🚀 Project Overview

Model: YOLOv8 Nano (custom-trained on Roboflow dataset).
Server: FastAPI (Python) hosting the model for inference.
Database: MongoDB Atlas (Vector Search for historical similarity).
Orchestration: n8n (Agentic workflow for decision making).

🛠 Prerequisites

Python 3.10+
n8n (Desktop or Self-hosted)
MongoDB Atlas Account (with Vector Search index enabled)
Optional: NVIDIA GPU with CUDA for faster training.

📦 Setup Instructions

1. Clone & Environment

git clone <your-repo-url>
cd palm-oil-ai
python -m venv venv
# Windows: venv\Scripts\activate | Mac: source venv/bin/activate
pip install -r requirements.txt

2. Dataset Preparation

Download the dataset from Roboflow Universe.
Unzip into the /datasets folder.
Ensure your data.yaml matches the local paths:
```
train: ../datasets/train/images
val: ../datasets/valid/images
```
3. Training the Model

To train locally without hanging your PC, use the throttled script:
```
python train_script.py
```

Outputs: The best model will be saved at runs/detect/train/weights/best.pt.
Move it: Copy best.pt to the root directory for the server to use.

4. Running the Inference Server

python main.py

The server will start at http://localhost:8000.

Endpoint: POST /detect
Payload: Multipart Form-data (Key: file, Value: image.jpg)

🤖 n8n Integration

The n8n workflow follows this logic:

Trigger: Receives image (Telegram/Webhook).
HTTP Request: Sends image to localhost:8000/detect.
MongoDB Node: Performs Vector Search using the returned embedding.
Agent Logic: Final ripeness determination based on model confidence + DB similarity.

📂 Repository Structure

├── datasets/           # Labeled images from Roboflow
├── runs/               # YOLO training logs and weights
├── main.py             # FastAPI Inference Server
├── train_script.py     # Local training configuration
├── best.pt             # The "Brain" (Trained Model)
├── requirements.txt    # dependencies
└── README.md           # You are here

📝 Future Improvements

Implement CLIP embeddings for higher-accuracy vector similarity.
Add a Streamlit dashboard for manual batch verification.

README.md 2.8 KB Historik Rå