Shirt Color Detection Web Application

By Saeed Anwar, Romail Afzal, Muhammad Masoom, Arslan Anwar

Contents

• Introduction
• Features
• Requirements
• Installation
• Usage
• Configuration

Introduction

This project is a Django-based web application for real-time Shirt color detection using YOLO v8. The application allows users to upload images, capture images from their camera, and perform real-time shirt color detection on the camera feed. The detected shirts are highlighted in the images using bounding boxes.

Features

• Upload and process images for shirt color detection.
• Capture images using your device's camera for real-time shirt color detection.
• Display the processed images with detected shirt color highlighted using bounding boxes.

Requirements

• Django
• torch
• numpy
• Pillow
• opencv-python
• ultralytics

Installation

1. Clone this repository to your local machine:git clone https://github.com/romailafzal/Shirt_color_detection_using-_yolov8

2. Create a virtual environment (optional but recommended):

   python3 -m venv venv source venv/bin/activate

3. Install the required dependencies: pip install -r requirements.txt

4. Run the Django development server: python manage.py runserver

5. Open your web browser and navigate to http://127.0.0.1:8000/ to access the application. 8000 is the port use your port accordingly

Usage

1. Upload Photo: Visit the upload page to select and upload an image from your local machine. The application will process the image and display it with the detected shirt color highlighted.

2. Camera Stream: Access the camera stream page to capture real-time images using your device's camera. The captured images will be processed for shirt color detection, and the results will be displayed.

3. Detected Image: View the processed image with detected shirt color highlighted.

4. Real-Time Detection: Use the real-time detection feature to stream live camera feed and see the detected shirt color in real time.

Configuration

• The YOLO model weights are loaded from the path /home/romail/ML1/trainning/SHIRT_DETECTION/v8best/best2.pt. You can update this path to point to your trained YOLO weights.

Note

• The application may require additional setup and configurations, especially related to camera access and YOLO model weights path.

Acknowledgments

This project was inspired by the need for a simple and interactive Shirt Color detection web application.