https://github.com/rishn/cognitive-robotics-pen-detection

https://github.com/rishn/cognitive-robotics-pen-detection

Last synced: 3 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/rishn/cognitive-robotics-pen-detection
• Owner: rishn
• Created: 2024-11-05T13:43:35.000Z (11 days ago)
• Default Branch: main
• Last Pushed: 2024-11-05T14:14:43.000Z (11 days ago)
• Last Synced: 2024-11-05T14:45:04.575Z (11 days ago)
• Language: Jupyter Notebook
• Size: 24.3 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Cognitive Robotics Pen Detection

## Aim

To build and evaluate an object detection system utilizing the YOLO (You Only Look Once) model for real-time object detection.

## Components

- **Dataset**: Images captured using a Raspberry Pi camera.

- **CNN Model**: YOLO model for object detection.

## Overview

This project leverages the YOLO model for detecting objects in images. The model is trained on a dataset created using a Raspberry Pi camera, capturing images of objects that are then annotated and processed for training.

## Procedure

1. **Image Capture**: Capture images using a Raspberry Pi camera and the `libcamera` module.

2. **Annotation and Dataset Preparation**: Use Roboflow to annotate the images and prepare the dataset.

3. **Model Training**: Utilize the YOLO model to train on the prepared dataset.

4. **Prediction**: Run the model on test images to detect and visualize objects.

## Running the Code

To execute the project, follow the steps outlined in the provided Jupyter notebook. You can also run the notebook in Google Colab for easy access and execution.

## Results

The YOLO model successfully detects objects, and the results are visualized with bounding boxes on images. The following performance metrics were observed:

- **Precision**: 0.80

- **Recall**: 0.67

- **Accuracy**: 0.57

### Confusion Matrix

|                         | Predicted: No Object | Predicted: Object |

|-------------------------|----------------------|--------------------|

| **Actual: No Object**   | 4                    | 15                |

| **Actual: Object**      | 18                   | 0                 |

### Example Outputs

Here are some examples of the model's predictions:







## Conclusion

This project demonstrates the capabilities of the YOLO model for real-time object detection, showcasing its effectiveness in visualizing detected objects with bounding boxes.