Pillow-SIMD is "following" Pillow fork (which is PIL fork itself).
There are many ways to improve the performance of image processing. You can use better algorithms for the same task, you can make better implementation for current algorithms, or you can use more processing unit resources. It is perfect when you can just use more efficient algorithm like when gaussian blur based on convolutions was replaced by sequential box filters. But a number of such improvements are very limited. It is also very tempting to use more processor unit resources (via parallelization) when they are available. But it is handier just to make things faster on the same resources. And that is where SIMD works better.
SIMD stands for "single instruction, multiple data". This is a way to perform same operations against the huge amount of homogeneous data. Modern CPU have different SIMD instructions sets like MMX, SSE-SSE4, AVX, AVX2, AVX512, NEON.
Currently, Pillow-SIMD can becompiledwith SSE4 (default) and AVX2 support.
Pillow-SIMD can be used in production. Pillow-SIMD has been operating on Uploadcare servers for more than 1 year. Uploadcare is SAAS for image storing and processing in the cloud and the main sponsor of Pillow-SIMD project.
Currently, following operations are accelerated:
- Resize (convolution-based resample): SSE4, AVX2
- Gaussian and box blur: SSE4
The numbers in the table represent processed megapixels of source image per second. For example, if resize of 7712×4352 image is done in 0.5 seconds, the result will be 61.7 Mpx/s.
- ImageMagick 6.9.3-8 Q8 x86_64
- Pillow 3.2.0
- Pillow-SIMD 3.2.0.post1
|Source||Operation||Filter||IM||Pillow||SIMD SSE4||SIMD AVX2|
|7712×4352||Resize to 16×16||Bilinear||27.0||217||456||545|
|Resize to 320×180||Bilinear||32.0||166||354||410|
|Resize to 2048×1155||Bilinear||20.7||87.6||202||217|
|1920×1280||Resize to 16×16||Bilinear||41.6||196||422||489|
|Resize to 320×180||Bilinear||27.6||111||196||197|
|Resize to 2048×1155||Bilinear||9.1||20.7||71.3||72.6|
Pillow is always faster than ImageMagick. And Pillow-SIMD is faster than Pillow in 2—2.5 times. In general, Pillow-SIMD with AVX2 almost always 10 times faster than ImageMagick.
All tests were performed on Ubuntu 14.04 64-bit running on Intel Core i5 4258U with AVX2 CPU on the single thread.
ImageMagick performance was measured with command-line tool
-bench arguments. I use command line because I need to test the latest version and this is the easiest way to do that.
All operations produce exactly the same results. Resizing filters compliance:
- PIL.Image.BILINEAR == Triangle
- PIL.Image.BICUBIC == Catrom
- PIL.Image.LANCZOS == Lanczos
In ImageMagick, the radius of gaussian blur is called sigma and the second parameter is called radius. In fact, there should not be additional parameters for gaussian blur , because if the radius is too small, this is not gaussian blur anymore. And if the radius is big this does not give any advantages but makes operation slower. For the test, I set the radius to sigma × 2.5.
Following script was used for testing: https://gist.github.com/homm/f9b8d8a84a57a7e51f9c2a5828e40e63
Why Pillow itself is so fast
There are no cheats. High-quality resize and blur methods are used for all benchmarks. Results are almost pixel-perfect. The difference is only effective algorithms. Resampling in Pillow was rewritten in version 2.7 with minimal usage of floating point numbers, precomputed coefficients and cache-awareness transposition.
Why Pillow-SIMD is even faster
Because of SIMD, of course. There are some ideas how to achieve even better performance.
- Efficient work with memory Currently, each pixel is read from memory to the SSE register, while every SSE register can handle four pixels at once.
- Integer-based arithmetic Experiments show that integer-based arithmetic does not affect the quality and increases the performance of non-SIMD code up to 50%.
- Aligned pixels allocation Well-known that the SIMD load and store commands work better with aligned memory.
Why do not contribute SIMD to the original Pillow
Well, it’s not that simple. First of all, Pillow supports a large number of architectures, not only x86. But even for x86 platforms, Pillow is often distributed via precompiled binaries. To integrate SIMD in precompiled binaries we need to do runtime checks of CPU capabilities. To compile the code with runtime checks we need to pass
-mavx2 option to the compiler. However this automatically activates all
if (__AVX2__) and below conditions. And SIMD instructions under such conditions exist even in standard C library and they do not have any runtime checks. Currently, I don’t know how to allow SIMD instructions in the code but do not allow such instructions without runtime checks.
In general, you need to do
pip install pillow-simd as always and if you are using SSE4-capable CPU everything should run smoothly. Do not forget to remove original Pillow package first.
If you want the AVX2-enabled version, you need to pass the additional flag to C compiler. The easiest way to do that is define
CC variable while compilation.
$ pip uninstall pillow $ CC="cc -mavx2" pip install -U --force-reinstall pillow-simd
Contributing to Pillow-SIMD
Pillow-SIMD and Pillow are two separate projects. Please submit bugs and improvements not related to SIMD tooriginal Pillow. All bugs and fixes in Pillow will appear in next Pillow-SIMD version automatically.
转载本站任何文章请注明：转载至神刀安全网，谢谢神刀安全网 » Pillow-SIMD is 2.5 times faster than Pillow and 10 times faster than ImageMagick