Pillow-SIMD is a performance-oriented fork of Pillow that accelerates selected image processing operations using SIMD instructions. SIMD, or Single Instruction Multiple Data, allows a processor to apply the same operation to multiple values simultaneously. The project is primarily designed for x86 processors and provides optimisations for workloads involving image resizing, blurring, colour conversion and alpha compositing.
The library follows Pillow’s programming interface, making it a replacement for corresponding Pillow functionality rather than a separate image-processing framework. Its optimisations are implemented in computationally intensive routines, where processing multiple pixels together can substantially reduce execution time. Pillow-SIMD supports SSE4 and AVX2 acceleration, with the actual performance depending on the processor, image dimensions and operation being performed. It is particularly relevant to applications processing many images or repeatedly resizing large photographs. Operations outside its accelerated code paths retain the general functionality inherited from Pillow.
This is free and open source software.
Key Features
- Pillow-compatible API – provides the familiar image objects and operations of corresponding Pillow implementations.
- SIMD acceleration – optimised processing paths using SSE4 and AVX2 instructions.
- Image resizing – accelerates convolution-based image resampling using SIMD operations.
- Gaussian blur – optimised Gaussian and box blur processing using SSE4.
- Alpha compositing – accelerates the combination of images containing transparency.
- Alpha premultiplication – optimises conversion from RGBA to premultiplied RGBa.
- Alpha unpremultiplication – accelerates conversion from premultiplied RGBa back to RGBA.
- Grayscale conversion – SIMD optimisations for RGB to luminance conversion.
- Convolution filters – accelerated 3×3 and 5×5 image kernels.
- Channel operations – optimised split and get_channel processing.
- Loop optimisations – uses loop unrolling and specialised instructions alongside SIMD processing.
- Inherited Pillow functionality – retains image format handling, drawing, transformations and other compatible image operations.
Website: github.com/uploadcare/pillow-simd
Support:
Developer: Uploadcare and Pillow contributors
License: MIT-CMU License
Pillow is written in Python and C. Learn Python with our recommended free books and free tutorials. Learn C with our recommended free books and free tutorials.
Related Software
| Image Processing Libraries | |
|---|---|
| matplotlib | Python 2D plotting library |
| OpenCV | Library that includes several hundreds of computer vision algorithms |
| VIPS | Fast image processing library with low memory requirements |
| SciPy | Scientific Computing Tools for Python |
| Pillow | Fork of the Python Imaging Library |
| Pillow-SIMD | Highly optimized downstream Pillow fork |
| scikit-image | Collection of algorithms for image processing |
| Kornia | Differentiable computer vision library |
| ImageMagick | Uses multiple computational threads to increase performance |
| GraphicsMagick | Billed as the Swiss army knife of image processing. |
| GEGL | Generic Graphics Library |
| OpenImageIO | Library and utilities for reading, writing and processing image files |
| CImg | Lightweight C++ toolkit for developing image-processing applications |
| Mahotas | Library of fast computer vision algorithms |
| SimpleITK | Image analysis toolkit with a large number of components |
| Netpbm | Toolkit for manipulation of graphic images |
| LibGD | Library for the dynamic creation of images by developers |
Read our verdict in the software roundup.
LinuxLinks helps you discover the best free and open source software for Linux. Explore our extensive collection of recommended applications, including alternatives to popular proprietary software.We also publish software reviews, Linux tutorials, hardware reviews, free programming books and articles covering the wider open source community. Browse our Linux resources to discover more. Know a useful free and open source Linux application we haven't covered? Suggest a program for inclusion. Has your software been featured on LinuxLinks? Display one of our Featured on LinuxLinks badges on your website or GitHub README. |


Please read our Comment Policy before commenting.