Edgar Riba

to reproduce: ```RUSTFLAGS="-C target-cpu=native" cargo bench read_jpeg``` image size HxWx3 -- 258x195 ; not big not small similar to what you might target to load large ml dataset in prod...

[feat] data exploration utils

2

- compute_std_mean - compute_histrogram - put examples to load a large directory and compute averaged std/mean and histograms - make_tiles (create a grid image) - show in rerun

fast image resize

3

Explore this crate and benchmark: https://docs.rs/fast_image_resize/latest/fast_image_resize/index.html

list of existing libraries - https://github.com/facontidavide/Bonxai

explore: https://github.com/elftausend/nvjpeg-rs

Explore the following to port in rust: - https://github.com/ducha-aiki/pydegensac @ducha-aiki - https://github.com/ivashmak/vsac (seems to be the fastest?) - https://github.com/PoseLib/PoseLib - https://github.com/danini/graph-cut-ransac/tree/master

reference codes: - https://github.com/kornia/kornia/pull/2658 - https://docs.opencv.org/4.x/dc/d6b/group__video__track.html#ga473e4b886d0bcc6b65831eb88ed93323 requirements: - gaussian pyramid: https://github.com/kornia/kornia/blob/main/kornia/geometry/transform/pyramid.py - gaussian blur https://github.com/kornia/kornia/blob/main/kornia/filters/gaussian.py#L13 - convolution filter https://github.com/kornia/kornia/blob/main/kornia/filters/filter.py#L38

explore/benchmark - https://github.com/image-rs/image-png - https://github.com/kornelski/lodepng-rust - https://crates.io/crates/zune-png

Support properly OpenEXR

3

the crate `image-rs` supports openexr images so that we can import/export floating point images. However, the `cv::Tensor` class assumes for now only byte representation.

[wip] add undistort image tutorial

1

undistorted image (for local testing while wip) @cjpurackal