SIGNED-DISTANCE-FIELD-RS

Fast Signed Distance Fields for Rust!

This crate approximates a signed distance field, given a binary image. The algorithm used is called "dead reckoning", as described in The "dead reckoning" signed distance transform by George J. Grevara (2004).

The complexity of the algorithm is linear, while an exact algorithm would have quadratic complexity.

On my machine, the distance field of an image with 4096 * 4096 (16 Megapixels) pixels can be computed in about 1.14 seconds.

Features

In the process of computing the signed distance field, the algorithm constructs an image with each pixel containing the vectors which point to the nearest edge. This vector distance field is made available after computing the plain distance field and can be used for further processing.

Piston Images

This library can be configured to offer some simple conversions to and from piston images. The feature flag piston_image unlocks these functions. The image crate is not required to calculate the signed distance field.

Getting Started

use sdf_dead_reckoning::prelude::*;
    
fn main(){
    let mut gray_image = image::open("images/sketch.jpg").unwrap().to_luma();
    let binary_image = binary_piston_image::of_gray_u8_image_with_threshold(&gray_image, 80);

    let distance_field = compute_f32_distance_field(&binary_image);
    let distance_image = distance_field.normalize_distances().to_gray_u8_image();

    distance_image.save("images/sketch_distance.png").unwrap();
}

Note: To run this specific example, use sdf-dead-reckoning with the piston image crate feature, by enabling the feature flag piston_image.

TODO

• Enable customized memory destination instead of predefined allocations
• Real Profiling and Benchmarking
• Consider SIMD and Multithread optimization?