A-KAZE Feature Detector, Extractor, and Matcher for Rust

This repository contains a Rust implementation of the A-KAZE algorithm. I would like to thank the original authors of the A-KAZE algorithm, who provided the implementation I used as a reference here:

A-KAZE original authors' GitHub repository

The work in this paper was first published in the following papers.

Fast Explicit Diffusion for Accelerated Features in Nonlinear Scale Spaces. Pablo F. Alcantarilla, J. Nuevo and Adrien Bartoli. In British Machine Vision Conference (BMVC), Bristol, UK, September 2013

KAZE Features. Pablo F. Alcantarilla, Adrien Bartoli and Andrew J. Davison. In European Conference on Computer Vision (ECCV), Fiorenze, Italy, October 2012

In case you found this elsewhere, this code lives here: https://github.com/indianajohn/akaze-rust/

Summary

The algorithm used here was heavily inspired by that repository, but differs in places. The resulting implementation produces results very similar to the original code, albeit a bit more slowly. The code in this crate is 100% safe Rust, as are most of the dependencies.

Results

This crate can be used to extract keypoints and descriptors from an image using a non-linear scale space.

It also includes a function to do matching with the 8-point algorithm.

Example

 extern crate akaze;
 use std::path::Path;
 let options = akaze::types::evolution::Config::default();
 let (_evolutions_0, keypoints_0, descriptors_0) =
     akaze::extract_features(
       Path::new("test-data/1.jpg").to_owned(),
       options);

 let (_evolutions_1, keypoints_1, descriptors_1) =
     akaze::extract_features(
       Path::new("test-data/1.jpg").to_owned(),
       options);
 let matches = akaze::match_features(&keypoints_0, &descriptors_0, &keypoints_1, &descriptors_1);
akaze::types::feature_match::serialize_to_file(&matches, Path::new("matches.cbor").to_owned());
println!("Got {} matches.", matches.len());

Running Demonstrations

# All executables (and your code probably) should be run in release mode, otherwise
# these can be quite slow.
# Extraction
cargo run --release --bin extract_features -- ./test-data/2.jpg ./output.cbor 

# Matching
cargo run --release --bin extract_and_match -- -m ./match_image.png ./test-data/1.jpg ./test-data/2.jpg

# Visualizing scale space
cargo run --release --bin extract_features -- ./test-data/2.jpg ./output.cbor  -d ./scale-space/

License

This code is released under the MIT license. See LICENSE.md for more details. You're free to use this however you'd like.

Disclaimer

I chose to do this project to learn Rust. It's entirely possible this code does not follow what some would consider the best Rust style. Use at your own risk.