Skip to content

Latest commit

 

History

History
144 lines (108 loc) · 4.43 KB

README.md

File metadata and controls

144 lines (108 loc) · 4.43 KB
Raw

Rust utilities for YUV format handling and conversion.

crates.io Build

Fast and simple YUV approximation conversion in pure Rust. At most the same as libyuv does. Performance will be equal to libyuv or slightly higher on platforms where SIMD is implemented. Otherwise equal or slower.

Mostly implemented AVX-512, AVX2, SSE, NEON, WASM

X86 targets with SSE and AVX uses runtime dispatch to detect available cpu features.

Supports:

  • YCbCr ( aka YUV )
  • YCgCo
  • YCgCo-R
  • YUY2
  • Identity ( GBR )
  • Sharp YUV

All the methods support RGB, BGR, BGRA and RGBA

SIMD

rustc avx2, avx512f, avx512bw, neon, sse4.1 features should be set when you expect than code will run on supported device.

For AVX-512 target feature avx512bw is required along with feature nightly_avx512 and nightly rust channel compiler.

Wasm simd128 should be enabled for implemented SIMD wasm paths support

Rayon

Some paths have multi-threading support, consider this feature if you're working on platform with multi-threading.

Adding to project

cargo add yuvutils-rs

RGB to YCbCr

let mut planar_image =
    YuvPlanarImageMut::<u8>::alloc(width as u32, height as u32, YuvChromaSubsampling::Yuv420);
rgb_to_yuv422(
    &mut planar_image,
    &src_bytes,
    rgba_stride as u32,
    YuvRange::Limited,
    YuvStandardMatrix::Bt601,
)
.unwrap();

YCbCr to RGB

yuv420_to_rgb(
    &yuv_planar_image,
    &mut rgba,
    rgba_stride as u32,
    YuvRange::Limited,
    YuvStandardMatrix::Bt601,
)
.unwrap();

Benchmarks

YUV 8 bit-depth conversion

cargo bench --bench yuv8 --manifest-path ./app/Cargo.toml

Tests performed on the image 5763x3842

Encoding 8-bit

time(NEON) Time(AVX)
utils RGB->YUV 4:2:0 3.16ms 3.53ms
libyuv RGB->YUV 4:2:0 3.58ms 33.87ms
utils RGBA->YUV 4:2:0 4.04ms 5.47ms
libyuv RGBA->YUV 4:2:0 4.87ms 23.48ms
utils RGBA->YUV 4:2:2 4.34ms 7.08ms
libyuv RGBA->YUV 4:2:2 5.90ms 35.23ms
utils RGBA->YUV 4:4:4 4.49ms 7.97ms

Decoding 8-bit

time(NEON) Time(AVX)
utils YUV NV12->RGB 3.86ms 6.24ms
libyuv YUV NV12->RGB 5.20ms 45.28ms
utils YUV 4:2:0->RGB 3.26ms 5.25ms
libyuv YUV 4:2:0->RGB 5.70ms 44.95ms
utils YUV 4:2:0->RGBA 3.77ms 5.98ms
libyuv YUV 4:2:0->RGBA 6.13ms 6.88ms
utils YUV 4:2:2->RGBA 4.88ms 6.91ms
libyuv YUV 4:2:2->RGBA 5.91ms 6.91ms
utils YUV 4:4:4->RGBA 4.79ms 7.20ms
libyuv YUV 4:4:4->RGBA 4.82ms 7.30ms

YUV 16 bit-depth conversion

cargo bench --bench yuv16 --manifest-path ./app/Cargo.toml

Encoding 10-bit

10-bit encoding is not implemented in libyuv

time(NEON) Time(AVX)
utils RGB10->YUV10 4:2:0 4.98ms 8.71ms
libyuv RGB10->YUV10 4:2:0 x x
utils RGBA10->YUV10 4:2:0 6.03ms 9.79ms
libyuv RGBA10->YUV10 4:2:0 x x
utils RGBA10->YUV10 4:2:2 5.99ms 13.54ms
libyuv RGBA10->YUV10 4:2:2 x x
utils RGBA10->YUV10 4:4:4 4.84ms 16.88ms

Decoding 10-bit

time(NEON) Time(AVX)
utils YUV10 4:2:0->RGB10 5.64ms 12.27ms
libyuv YUV10 4:2:0->RGB10 - -
utils YUV10 4:2:0->RGBA10 6.26ms 14.79ms
utils YUV10 4:2:0->RGBA8 7.05ms 42.44ms
libyuv YUV10 4:2:0->RGBA8 12.79ms 63.02ms
utils YUV10 4:2:2->RGBA10 6.31ms 15.33ms
utils YUV10 4:2:2->RGBA8 7.33ms 13.55ms
libyuv YUV10 4:2:2->RGBA10 12.62ms 58.98ms
utils YUV10 4:4:4->RGBA10 6.39ms 16.39ms

This project is licensed under either of

  • BSD-3-Clause License (see LICENSE)
  • Apache License, Version 2.0 (see LICENSE)

at your option.