https://github.com/xeptore/grayscaler-cuda

Simple educational-purpose multi-threaded image grayscaler - CUDA version
https://github.com/xeptore/grayscaler-cuda

Last synced: 1 day ago
JSON representation

Simple educational-purpose multi-threaded image grayscaler - CUDA version

• Host: GitHub
• URL: https://github.com/xeptore/grayscaler-cuda
• Owner: xeptore
• License: agpl-3.0
• Created: 2021-07-11T03:28:20.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2021-07-13T14:25:43.000Z (over 3 years ago)
• Last Synced: 2024-11-16T09:13:07.733Z (2 months ago)
• Language: Cuda
• Size: 1.65 MB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# CUDA Grayscaler

Simple educational-purpose multi-threaded image grayscaler - CUDA version

Currently, only jpeg images are supported using `libjpeg`.

## Usage

1. Configure

   You can configure

      - input image name (`INPUT_IMAGE_FILENAME`)

      - output images name (`OUTPUT_IMAGE_FILENAME`)

      - number of threads in each block (`BLOCK_THREADS`)

      - number of blocks in each grid (`GRID_BLOCKS`)

   configuration parameters in [`config.h`](/config.h) file.

2. Build

   ```sh

   make build

   ```

3. Run

   ```sh

   ./build/grayscale

   ```

## Development

You'll need CMake and a C compiler. I used CMake version `3.20.5` and clang version `12.0.1`.