Object Detection Using LLM

This project combines Large Language Models (LLMs) and YOLOv8 for object detection and identification in images. It uses Ollama for vision-based part identification and Ultralytics YOLO for object detection and annotation.

Features

• Identifies visible parts in an image using Ollama LLM (vision model)
• Detects and annotates objects in the image using YOLOv8
• Saves the annotated image with bounding boxes for matched parts

Requirements

• Python 3.8+
• ollama
• ultralytics
• opencv-python

Installation

1. Clone this repository or download the code.
2. Install dependencies:

   pip install -r requirements.txt

3. Ensure you have the YOLOv8 model weights (e.g., yolov8m.pt).
4. Place your input image at the specified path in main.py.

Usage

Edit the image_path and output_path in main.py as needed, then run:

python main.py

Notes

• Ollama must be running and accessible for the LLM vision step.
• The list of parts to identify can be customized in the prompt in main.py.
• The YOLO model and weights can be changed as needed.

Output

• The script prints the detected parts and saves an annotated image to the specified output path.