imgproc-rs

A Rust image processing library.

Features

• Multithreading support for some functions via rayon
• SIMD support for some functions using AVX2

Supported Image Formats

imgproc-rs uses the i/o functions provided in the image crate. A list of supported image formats can be found here.

Examples

Reading and Writing Images

use imgproc_rs::io;

fn main() {
    // Read an image from a path
    let img = io::read("path/to/some_image.png").unwrap();
    
    // Print image information
    println!("{:?}", img.info());
    
    // Write the image to a path as a PNG
    io::write(&img, "path/to/save_image.png").unwrap();
}

Creating Images

Images can be created from existing vectors, slices, vectors of vectors, and vectors of slices. A few examples are shown below.

use imgproc_rs::image::{Image, ImageInfo};

fn main() {
    let vec = vec![1, 2, 3, 4, 5, 6,
                   7, 8, 9, 10, 11, 12];

    // Create an image from a slice
    let img_slice = Image::from_slice(2, 2, 3, false, &vec);

    // Create an image from a vector
    let img_vec = Image::from_vec(2, 2, 3, false, vec);

    // Create a blank (black) image
    let img_blank: Image<u8> = Image::blank(ImageInfo::new(2, 2, 3, false));

    // Create an empty image
    let img_empty: Image<u8> = Image::empty(ImageInfo::new(2, 2, 3, false));
}

Getting Image Information

use imgproc_rs::io;
use imgproc_rs::image::{Image, ImageInfo};

fn main() {
    let img = io::read("path/to/some_image.png").unwrap();

    // Get width and height of image
    let (width, height) = img.info().wh();

    // Get width, height, and channels of image
    let (width, height, channels) = img.info().whc();

    // Get width, height, channels, and alpha of image
    let (width, height, channels, alpha) = img.info().whca();

    /* Print image information
     * Example output:
     *
     * width: 2
     * height: 2
     * channels: 3
     * alpha: false
     *
     */
    println!("{}", img.info());
}

Getting/Setting Image Pixels

Image pixels can be accessed using either 1D or 2D indexing. 1D indexing reads the image data row by row from left to right, starting in the upper left corner of the image. 2D coordinates start at zero in the upper left corner of the image and increase downwards and to the right.

use imgproc_rs::io;
use imgproc_rs::image::{Image, BaseImage};

fn main() {
    let img = io::read("path/to/some_image.png").unwrap();
    
    // Set an image pixel using a 1D index
    img.set_pixel_indexed(0, &[1, 1, 1]);
    
    // Get an image pixel using a 1D index
    let pixel_1d = &img[0];
    
    // Set an image pixel using 2D coordinates
    img.set_pixel(1, 1, &[1, 1, 1]);
    
    // Get an image pixel using 2D coordinates
    let pixel_2d = img.get_pixel(1, 1);
}

Multithreading

To enable multithreading, include the parallel feature in your Cargo.toml:

[dependencies.imgproc-rs]
version = "0.3.0"
default-features = false
features = ["parallel"]

Alternatively, pass the features flag to cargo run:

cargo run --features parallel

Functions that support multithreading:

• transform module
  • crop
  • scale
  • scale_lanczos
• All functions in the filter module, except:
  • threshold
  • residual
  • median_filter
  • alpha_trimmed_mean_filter

SIMD

SIMD support is enabled by default via the simd feature. Supported functions will only use SIMD where AVX2 is supported (x86/x86_64); otherwise an alternate implementation is used.

Functions that support SIMD (outside simd module)

• tone::brightness()
• tone::saturation()
• colorspace::rgb_to_grayscale()