Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/Alyetama/Ray-Image
Fast image compression for large number of images with Ray library
https://github.com/Alyetama/Ray-Image
compression image-compression jpeg preprocessing python ray
Last synced: 3 months ago
JSON representation
Fast image compression for large number of images with Ray library
- Host: GitHub
- URL: https://github.com/Alyetama/Ray-Image
- Owner: Alyetama
- License: mit
- Created: 2022-04-28T20:42:33.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-12-06T17:06:28.000Z (11 months ago)
- Last Synced: 2024-05-06T23:01:29.263Z (6 months ago)
- Topics: compression, image-compression, jpeg, preprocessing, python, ray
- Language: Python
- Homepage: https://alyetama.github.io/Ray-Image
- Size: 47.9 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Ray-Image
🚀 Fast image compression for large number of images with Ray library.
[](https://www.python.org/downloads/) [](https://www.python.org/dev/peps/pep-0008/) [](https://github.com/Alyetama/Ray-Image/actions/workflows/poetry-build.yml)
## Requirements
- 🐍 [Python>=3.7](https://www.python.org/downloads/)
- ⚡ [Ray>=1.0.0](https://github.com/ray-project/ray)
To install `ray`, run\*:
```
pip install ray
```
\*For Apple Silicon (M1), follow the instructions [here](https://docs.ray.io/en/latest/ray-overview/installation.html#m1-mac-apple-silicon-support) to install `ray`.
## ⬇️ Installation
```
pip install rayim
```
## ⌨️ Usage
```
usage: rayim [-h] [-o OUTPUT_DIR] [-q QUALITY] [--overwrite] [-n] [-j]
[--replicate-dir-tree] [-s SIZE SIZE] [-d DIV_BY] [-S] [-O]
path [path ...]
positional arguments:
path Path to a single file/directory or multiple
files/directories
options:
-h, --help show this help message and exit
-o OUTPUT_DIR, --output-dir OUTPUT_DIR
Output directory (default: next to original file)
-q QUALITY, --quality QUALITY
Output image quality (JPEG only; default: 70)
--overwrite Overwrite the original image
-n, --no-subsampling Turn off subsampling and retain the original image
setting (JPEG only)
-j, --to-jpeg Convert the image(s) to .JPEG
--replicate-dir-tree Replicate the source directory tree in the output
-s SIZE SIZE, --size SIZE SIZE
Resize the image to WIDTH HEIGHT
-d DIV_BY, --div-by DIV_BY
Divide the image size (WxH) by a factor of n
-S, --silent Silent mode
-O, --optimize Apply default optimization on the image(s)
-k, --keep-date Keep the original modification date
```
## 📕 Examples
- Running on a single file:
```shell
rayim foo.jpg
# 🚀 foo.jpg: 1157. kB ==> 619.9 kB (-46.4%) | 0.07s
```
- Running on a folder `foo` and writing the output to `compressed`
```shell
rayim foo/ -o compressed
# (compress_many pid=612778) 🚀 foo.jpg: 988.9 kB ==> 544.8 kB (-44.9%) | 0.08s
# (compress_many pid=612828) 🚀 bar.jpg: 983.7 kB ==> 541.2 kB (-44.9%) | 0.07s
# (compress_many pid=612826) 🚀 foobar.jpg: 1001. kB ==> 550.7 kB (-44.9%) | 0.07s
# (compress_many pid=612786) 🚀 barfoo.jpg: 1001. kB ==> 551.9 kB (-44.8%) | 0.08s
# ...
# Total:
# Before: 1091.32 MB
# After: 599.46 MB (-45.0%)
```
# Speed comparison
### Test 1 (on Apple Silicon M1, 8-cores)
| Method | Number of files | Speed |
| ----------- | ----------- | ----------- |
| Regular compression | 1,000 | `60.090s` |
| rayim | 1,000 | **`26.937s`** (**`55.17%`** faster) |
```YAML
Total:
Before: 1091.32 MB
After: 599.46 MB (-45.0%)
```
### Test 2 (on Intel @ 2.299GHz, 32-cores)
| Method | Number of files | Speed |
| ----------- | ----------- | ----------- |
| Regular compression | 6,000 | `7m42.919s` |
| rayim | 6,000 | **`5m15.423s`** (**`31.96%`** faster) |
```YAML
Total:
Before: 6040.59 MB
After: 3321.70 MB (-45.0%)
```