Closed arthmis closed 8 months ago
I wondered about this too. AFAIK, iterating over pixels in parallel currently requires ugly code (uglier than a Rayon iterator, at least).
Is it necessary to add multi-threading? I was thinking about how it would benefit the imageproc crate, at least on large images, but maybe it would be better handled by the GPU.
i created this https://github.com/zzwxh/parallel_image/blob/main/src/lib.rs
Would adding multi-threading or parallel iterators to access pixels be feasible? Is it potentially planned for the future?