Skip to content

About

A computer vision project that detects and identifies multiple objects in images using deep learning and object detection techniques.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

9 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 

Repository files navigation

Object-detectionCV

Open In Colab YOLOv8 Nano MS COCO 80 Classes OpenCV License


⚡ Overview

Object-detectionCV is an end-to-end, real-time multi-category object detection pipeline powered by Ultralytics YOLOv8.

Trained on the 80 common object classes of the Microsoft Common Objects in Context (MS COCO) dataset, this application detects, classifies, and annotates multiple simultaneous objects—including vehicles, pedestrians, animals, furniture, and electronic devices—in a single forward pass.


🚀 Key Features

  • 🧠 State-of-the-Art Architecture: Utilizes Ultralytics YOLOv8 Nano (yolov8n.pt) with an anchor-free split head for faster inference and superior accuracy.
  • 🌐 80 MS COCO Object Classes: Recognizes a vast spectrum of everyday objects (cars, persons, dogs, traffic lights, laptops, cell phones, etc.).
  • 🎨 Built-in Color-Coded Annotation: Employs results[0].plot() for automated, high-precision bounding boxes with unique per-class color coding and confidence percentages.
  • ⚡ Ultra-Low Latency: High-speed inference (~12ms on GPU) suitable for robotics, drone perception, and security feeds.
  • ☁️ 1-Click Google Colab Workflow: Interactive file uploader with OpenCV rendering using cv2_imshow().

🎮 Workflow & Detection Pipeline

flowchart TD
    A["Input Image (Colab Upload)"] --> B["OpenCV Image Loading: cv2.imread()"]
    B --> C["YOLOv8n Deep Neural Network Inference: model(img)"]
    
    subgraph YOLOv8_Multi_Class_Engine["Ultralytics Detection Core"]
        C --> D["Feature Backbone & PAN-FPN Neck"]
        D --> E["Anchor-Free Bounding Box Prediction"]
        D --> F["80-Class Softmax Probability Estimation"]
        E --> G["Non-Maximum Suppression (NMS) Filtering"]
        F --> G
    end
    
    G --> H["results[0].plot() Automated Visual Rendering"]
    H --> I["Generate Color-Coded Bounding Boxes & Confidence Tags"]
    I --> J["Display Rendered Frame via cv2_imshow(output)"]
Loading

🏷️ Supported Object Categories (MS COCO)

YOLOv8 can identify all 80 classes across major categories:

Category Typical Detected Classes
People & Accessories person, backpack, umbrella, handbag, tie, suitcase
Vehicles & Transport bicycle, car, motorcycle, airplane, bus, train, truck, boat
Traffic & Outdoor traffic light, fire hydrant, stop sign, parking meter, bench
Animals bird, cat, dog, horse, sheep, cow, elephant, bear, zebra, giraffe
Electronics & Office tv, laptop, mouse, remote, keyboard, cell phone, book, clock
Food & Dining bottle, wine glass, cup, fork, knife, spoon, bowl, banana, apple, pizza

💻 How to Run in Google Colab

  1. Launch in Colab:
    Click the "Open In Colab" badge above or open in Google Colab.

  2. Execute Cell:
    The script automatically installs ultralytics and downloads the lightweight yolov8n.pt weights.

  3. Upload an Image:
    Upload any scene containing one or many objects (street, office, park, living room).

  4. Instant Annotation:
    YOLOv8 annotates every detected object with distinct color-coded bounding boxes, class names, and confidence scores.


🌐 Running Locally

To run the project on your local machine:

  1. Clone the repository:

    git clone https://github.com/dipzz0/Object-detectionCV.git
    cd Object-detectionCV
  2. Install dependencies:

    pip install ultralytics opencv-python
  3. Run the script:

    python Object-detection.py

    (Note: Replace cv2_imshow(output) with cv2.imshow('Detections', output) and cv2.waitKey(0) when running locally).


🛠️ Technology Stack

Technology Purpose Description
Ultralytics YOLOv8 Model & Inference Real-time object detection and segmentation framework
PyTorch Deep Learning Backend PyTorch-powered tensor compute engine
OpenCV Vision & Display Image I/O and display pipeline
MS COCO Training Benchmark Standard dataset with 80 annotated categories
Google Colab Execution Environment Cloud GPU/CPU serverless platform

📄 License

This repository is distributed under the MIT License.


Developed with ❤️ by dipzz0

About

A computer vision project that detects and identifies multiple objects in images using deep learning and object detection techniques.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages