You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We have mapwindow that applies function f on every local patch of an image, which can be seen as the image version of map. We don't have the window version of findall that applies f to each local patch, and return coordinates such that f(window_of_p) == true.
Why we need this? We can currently do this by returning a dense Bool matrix from mapwindow and then call findall, but this would usually be unnecessary and not as performant as the manually written for-loops. For example, the benchmark in JuliaImages/ImageSegmentation.jl#72 (comment)
The text was updated successfully, but these errors were encountered:
Originally posted in #223 (review)
We have
mapwindow
that applies functionf
on every local patch of an image, which can be seen as the image version ofmap
. We don't have the window version offindall
that appliesf
to each local patch, and return coordinates such thatf(window_of_p) == true
.Why we need this? We can currently do this by returning a dense Bool matrix from
mapwindow
and then callfindall
, but this would usually be unnecessary and not as performant as the manually written for-loops. For example, the benchmark in JuliaImages/ImageSegmentation.jl#72 (comment)The text was updated successfully, but these errors were encountered: