Function references

jpeg_encode(filename::AbstractString, img; kwargs...) -> Int
jpeg_encode(io::IO, img; kwargs...) -> Int
jpeg_encode(img; kwargs...) -> Vector{UInt8}

Encode 2D image img as JPEG byte sequences and write to given I/O stream or file. The return value is number of bytes. If output is not specified, the encoded result is stored in memory as return value.

Parameters

• transpose::Bool: whether we need to permute the image's width and height dimension before encoding. The default value is false.
• quality::Int: Constructs JPEG quantization tables appropriate for the indicated quality setting. The quality value is expressed on the 0..100 scale recommended by IJG. The default value is 92. Pass quality=nothing to let libjpeg-turbo dynamicly guess a value.

Examples

julia> using JpegTurbo, TestImages

julia> img = testimage("cameraman");

julia> jpeg_encode("out.jpg", img) # write to file
51396

julia> buf = jpeg_encode(img); length(buf) # directly write to memory
51396

References

• [1] libjpeg API Documentation (libjpeg.txt)

jpeg_decode([T,] filename::AbstractString; kwargs...) -> Matrix{T}
jpeg_decode([T,] io::IO; kwargs...) -> Matrix{T}
jpeg_decode([T,] data::Vector{UInt8}; kwargs...) -> Matrix{T}

Decode the JPEG image as colorant matrix. The source data can be either a filename, an IO , or an in-memory byte sequence.

parameters

• transpose::Bool: whether we need to permute the image's width and height dimension before encoding. The default value is false.
• scale_ratio::Real: scale the image by ratio scale_ratio in M/8 with M ∈ 1:16. The default value is 1. For values are not in the range, they will be mapped to the nearest value, e.g., 0.3 => 2/8 and 0.35 => 3/8. scale_ratio and preferred_size may not be used together.
• preferred_size::Tuple: infer the minimal scale_ratio that all(size(out) .>= preferred_size)) holds. It can optionally be (op, preferred_size) format, with compare operation op be one of >, >=, < or <=. If op in (>=, >) then it gets the minimal scale_ratio, otherwise it gets the maximum scale_ratio for op in (<=, <). scale_ratio and preferred_size may not be used together. The preferred_size dimensions are not affected by keyword transpose.

Examples

julia> using JpegTurbo, TestImages, ImageCore

julia> filename = testimage("earth", download_only=true);

julia> img = jpeg_decode(filename); summary(img)
"3002×3000 Array{RGB{N0f8},2} with eltype RGB{N0f8}"

julia> img = jpeg_decode(Gray, filename; scale_ratio=0.25); summary(img)
"751×750 Array{Gray{N0f8},2} with eltype Gray{N0f8}"

For image preview and similar purposes, T and scale_ratio are useful parameters to accelerate the JPEG decoding process. For color JPEG image, jpeg_decode(Gray, filename) is faster than jpeg_decode(filename) since the color components need not be processed. Smaller scale_ratio permits significantly faster decoding since fewer pixels need be processed and a simpler IDCT method can be used.

using BenchmarkTools, TestImages, JpegTurbo
filename = testimage("earth", download_only=true)
# full decompression
@btime jpeg_decode(filename); # 224.760 ms (7 allocations: 51.54 MiB)
# only decompress luminance component
@btime jpeg_decode(Gray, filename); # 91.157 ms (6 allocations: 17.18 MiB)
# process only a few pixels
@btime jpeg_decode(filename; scale_ratio=0.25); # 77.254 ms (8 allocations: 3.23 MiB)
# process only a few pixels for luminance component
@btime jpeg_decode(Gray, filename; scale_ratio=0.25); # 63.119 ms (6 allocations: 1.08 MiB)

versioninfo(io=stdout)

Print information about the libjpeg-turbo in use.