Repository files navigation

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. Waiting for SVE to be available innightly.

x86 targets with SSE and AVX uses runtime dispatch to detect available cpu features.

Supports:

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

All the methods support RGB, BGR, BGRA and RGBA

If need scaling support thenpic-scale is available.

Runtime dispatch is used for use if availablesse4.1,avx2,avx512bw,avx512vbmi,avx512vnni,avxvnni,rdm for NEON.

For thesse4.1,avx2,rdm there is no action needed, however for AVX-512 activated featurenightly_avx512 alongnightly rust channel is required.

Always consider to compile withnightly_avx512 when compiling for x86, it adds noticeable gain on supported cpus.

Wasmsimd128 as target feature should be enabled for implemented SIMD wasm paths support,

Some paths have multi-threading support. However, YUV conversion usually makes more sense in single-threaded mode, thus use this feature with care.

cargo add yuvutils-rs

letmut planar_image =YuvPlanarImageMut::<u8>::alloc(widthasu32, heightasu32,YuvChromaSubsampling::Yuv420);rgb_to_yuv422(&mut planar_image,&src_bytes,    rgba_strideasu32,YuvRange::Limited,YuvStandardMatrix::Bt601,).unwrap();

yuv420_to_rgb(&yuv_planar_image,&mut rgba,    rgba_strideasu32,YuvRange::Limited,YuvStandardMatrix::Bt601,).unwrap();

The following planar formats are used for HDR content, with either10-bit or12-bit color depth:

• I010, I012, P010, P012, P016
  Half width, half height
• I210, I212, P210, P212, P216
  Half width, full height
• I410, I412, P410, P412
  Full width, full height

• I: Represents the color space (as defined above), and these formats use3 planes: Y (luminance), U (chrominance), and V (chrominance).
• ICgC: Represents the color space (as defined above), and these formats use3 planes: Y (luminance), Cg (chrominance), and Co (chrominance).
• P: Abiplanar format (similar toNV12), but with16-bit precision, where the valid bits are stored in the high bits. This format has:
  • AY plane
  • AUV plane (with U and V interleaved)

• 0: Represents4:2:0 chroma subsampling.
• 2: Represents4:2:2 chroma subsampling.
• 4: Represents4:4:4 chroma subsampling.

• 10,12,14,16: Indicates thebits per channel.

Tests performed on the image 1997x1331

YUV 8 bit-depth conversion

aarch64 tested on Mac Pro M3.

AVX2 tests performed onm4.large with Ubuntu 24.04 droplet.AVX2 Win test performed on Windows 11 Intel Core i9-14900HX.

cargo bench --bench yuv8 --manifest-path ./app/Cargo.toml --features fast_mode,professional_mode

AVX-512 tests performed onAWS c5.large with Ubuntu 24.04 instance with command

cargo +nightly bench --bench yuv8 --manifest-path ./app/Cargo.toml --features nightly_avx512,fast_mode,professional_mode

YUV 16 bit-depth conversion

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

AVX-512 tests performed onAWS c5.large with Ubuntu 24.04 instance with command

cargo +nightly bench --bench yuv16 --manifest-path ./app/Cargo.toml --features nightly_avx512

10-bit encoding is not implemented inlibyuv

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

This project is licensed under either of

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

at your option.