https://github.com/yashkathe/image-noise-reduction-with-cuda

This repository conducts a comprehensive analysis of image denoising technique - median blur, comparing GPU-accelerated (Numba) and CPU-based (OpenCV) processing speeds. Using diverse images, the project applies median filtering to assess efficiency providing insights into the practical impacts of hardware acceleration in real-world applications
https://github.com/yashkathe/image-noise-reduction-with-cuda

cuda cuda-programming gpu-programming hardware-speed-analysis image-analysis image-processing numba nvidia nvidia-cuda nvidia-gpu opencv parallel-programming

Last synced: about 2 months ago
JSON representation

This repository conducts a comprehensive analysis of image denoising technique - median blur, comparing GPU-accelerated (Numba) and CPU-based (OpenCV) processing speeds. Using diverse images, the project applies median filtering to assess efficiency providing insights into the practical impacts of hardware acceleration in real-world applications

• Host: GitHub
• URL: https://github.com/yashkathe/image-noise-reduction-with-cuda
• Owner: yashkathe
• License: mit
• Created: 2023-11-24T20:00:32.000Z (about 1 year ago)
• Default Branch: master
• Last Pushed: 2024-03-17T21:14:19.000Z (9 months ago)
• Last Synced: 2024-03-17T22:27:54.967Z (9 months ago)
• Topics: cuda, cuda-programming, gpu-programming, hardware-speed-analysis, image-analysis, image-processing, numba, nvidia, nvidia-cuda, nvidia-gpu, opencv, parallel-programming
• Language: Jupyter Notebook
• Homepage:
• Size: 24.8 MB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# GPU vs CPU Runtime Analysis

## Overview

This repository contains a project that compares the performance of image

processing operations when executed on a GPU vs. a CPU. The focus is on

analyzing the execution time for median filtering across a set of images,

providing insights into the efficiency gains achievable with GPU acceleration.

## Project Description

The project uses Python, with OpenCV for CPU-based image processing and Numba

for GPU acceleration. The primary goal is to measure and compare the execution

time for median filtering—an image denoising technique—on both the CPU and GPU.

## Key Features

- Image Processing: Applies median filtering to a set of images using both CPU and GPU.

- Runtime Comparison: Measures and logs the execution time for both methods.

- Visualization: Displays original and processed images side by side for visual comparison.

## Getting Started

### Prerequisites

- Python 3.x

- Libraries: OpenCV, Numba, Matplotlib, NumPy

- CUDA-enabled GPU for running GPU-accelerated code

### Installation

Clone this repository or download the source code.  

Install required Python packages:

```bash

pip install opencv-python numba matplotlib numpy

```

### Usage

Place your images in the image-data directory.  

Run the Jupyter Notebook.  

The notebook will process the images and display the results along with the

runtimes.