Skip to content

Latest commit

 

History

History
110 lines (72 loc) · 3.35 KB

Readme.md

File metadata and controls

110 lines (72 loc) · 3.35 KB

Spatial Filtering Dashboard in Image Processing

An interactive Google Colab dashboard to demonstrate smoothing and sharpening filters for image processing.
Includes Mean, Median, and Mode smoothing filters, along with Sobel, Laplacian, and Sobel+Laplacian sharpening filters.
Useful for noise reduction, edge detection, and detail enhancement in real-world applications.

🔍 Introduction

Spatial filtering is an essential step in many computer vision and image processing tasks.
It helps in:

  • Enhancing details
  • Detecting edges
  • Reducing noise
    This project demonstrates how different filters behave and how smoothing can be combined with sharpening to improve results.

📂 Dataset

You can use:

  • Any grayscale or color image uploaded through the dashboard.
  • Recommended: datasets where edges and details matter, such as medical images, satellite images, or defect detection samples.

⚙️ Methodology & Justification

The following filters are implemented:

Smoothing Filters

  • Mean Filter: Averages local pixels, reduces Gaussian-like noise.
  • Median Filter: Non-linear filter effective against salt-and-pepper noise.
  • Mode Filter: Replaces pixels with the most frequent neighbor value, good for impulse noise.

Sharpening Filters

  • Sobel Filter (First Derivative): Highlights intensity gradients and edges.
  • Laplacian Filter (Second Derivative): Captures fine details, sensitive to noise.
  • Sobel + Laplacian: Combines directional edge detection with second-order enhancement.

Combined Pipeline

Smoothing → Sharpening
Prevents noise amplification, produces clearer edges.

🚀 Setup

Clone the repo and open in Google Colab:

git clone https://github.com/CodeRafay/Spatial-Filtering-Image-Processing.git
cd spatial-filtering-image-processing

Then open the notebook in Colab and run all cells. Required libraries:

pip install opencv-python-headless matplotlib ipywidgets scipy

📊 Usage

  1. Upload an image from your local machine.
  2. Choose a smoothing filter (Mean, Median, Mode).
  3. Choose a sharpening filter (Sobel, Laplacian, Sobel+Laplacian).
  4. Optionally enable "Apply smoothing before sharpening" for combined results.
  5. Compare original vs processed images, histograms, and filter explanations.

📈 Results & Analysis

  • Mean filter: Reduces noise but blurs edges.
  • Median filter: Preserves edges while removing salt-and-pepper noise.
  • Mode filter: Stabilizes impulse noise but less common in practice.
  • Sobel: Good for detecting strong directional edges.
  • Laplacian: Captures fine details but amplifies noise.
  • Combined smoothing + sharpening: Produces balanced results, especially with Median + Sobel.

📜 Citation

If you use this project in research or coursework, please cite:

@software{spatial_filtering_dashboard,
  author = Rafay Adeel,
  title = {Spatial Filtering Dashboard in Image Processing},
  year = {2025},
  url = {https://github.com/CodeRafay/Spatial-Filtering-Image-Processing}
}

🏷️ Keywords

Image Processing, Spatial Filters, Smoothing, Sharpening, Sobel, Laplacian, Median Filter, Edge Detection, Colab Dashboard