Skip to content

Object Detection using YOLOv8 A custom object detection model was developed using YOLOv8. The dataset was collected and annotated with LabelImg, then trained and fine-tuned with data augmentation to improve accuracy. The model was evaluated using mAP, precision, and recall, and deployed for real-time detection through OpenCV with a webcam feed.

Notifications You must be signed in to change notification settings

123sowji/Object-Detection-using-Python-with-YOLOv8

Repository files navigation

Overview

This project offers a graphical user interface (GUI) for real-time object detection using the YOLOv8 model. Users can:

Upload images from their local system for object detection.

Utilize their webcam for real-time object detection.

The application leverages the YOLOv8 model from the Ultralytics library, ensuring state-of-the-art performance in object detection tasks.

Features

Image Upload: Select and process images from your local storage to detect objects.

Webcam Integration: Activate your webcam to perform real-time object detection.

Bounding Boxes: Visualize detected objects with bounding boxes, labels, and confidence scores.

Installation

Prerequisites

Ensure you have the following installed:

Python 3.8 or higher

pip (Python package installer)

Using the Application

Upload Image:

Click the "Upload Image" button.

Select an image file from your local system.

The application will display the image with detected objects highlighted.

Start Webcam:

Click the "Start Webcam" button.

A new window will display the webcam feed with real-time object detection.

Press the 'ctrl+c' key to exit the webcam feed.

Acknowledgements

Ultralytics: For the YOLOv8 model and related resources.

OpenCV: For real-time computer vision capabilities.

Tkinter: For the GUI components.

About

Object Detection using YOLOv8 A custom object detection model was developed using YOLOv8. The dataset was collected and annotated with LabelImg, then trained and fine-tuned with data augmentation to improve accuracy. The model was evaluated using mAP, precision, and recall, and deployed for real-time detection through OpenCV with a webcam feed.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages