https://github.com/trflorian/image-processing

Explore multiprocessing for an image processing task
https://github.com/trflorian/image-processing

data-science opencv-python python

Last synced: about 1 year ago
JSON representation

Explore multiprocessing for an image processing task

• Host: GitHub
• URL: https://github.com/trflorian/image-processing
• Owner: trflorian
• Created: 2024-12-08T13:40:19.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2025-01-19T09:53:38.000Z (over 1 year ago)
• Last Synced: 2025-04-15T05:17:43.398Z (about 1 year ago)
• Topics: data-science, opencv-python, python
• Language: Python
• Homepage: https://medium.com/towards-data-science/how-to-process-10k-images-in-seconds-e99661f5c2b5
• Size: 55.7 KB
• Stars: 4
• Watchers: 1
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Image Proceesing

A small image processing project exploring how to use multiprocessing to optimize a data procesisng pipeline by parallelizing a CPU bound task on multiple cores.

![Comparison of processing time for different workers](timing.png)

## Quickstart

To get started, simply clone the repo, make sure that you have uv installed and then snyc the project dependencies.

```

uv sync

```

Place some images in `./images/` folder or specify a custom input image directory with the `--input_path` argument. Then run the program, it will show a progress bar in the terminal indicating the progress and the iteration speed. 

```

uv run main.py

```

By default the program will run on a single core. If you specify e.g. `--num_workers 10` it will use a pool of 10 processes and parallelize the task across them. 

## Usage

```

usage: main.py [-h] [--input_path INPUT_PATH] [--output_path OUTPUT_PATH] [--image_extensions IMAGE_EXTENSIONS [IMAGE_EXTENSIONS ...]]

               [--num_workers NUM_WORKERS] [--timing] [--subset SUBSET]

options:

  -h, --help            show this help message and exit

  --input_path INPUT_PATH

                        Path to a folder with images

  --output_path OUTPUT_PATH

                        Path where the processed images will be stored

  --image_extensions IMAGE_EXTENSIONS [IMAGE_EXTENSIONS ...]

                        Extensions of image files to look for

  --num_workers NUM_WORKERS

                        Specify number of workers to use. If not specified, will use a single process for loop.

  --timing              Enable timing of the process and save results to a `timing.txt` file

  --subset SUBSET       Specify a subset of the images to process

```