pub fn canny_into(
input: ImageView<'_, u8, 1>,
options: CannyOptions,
output: ImageViewMut<'_, u8, 1>,
workspace: &mut CannyWorkspace,
) -> VisionResult<()>Expand description
Finds edges into caller-owned output using reusable CPU workspace.
The output may be packed or strided, must match the input dimensions, and
receives only binary values (0 or 255). The specialized allocation-light
path applies to the common 3x3 aperture; larger apertures retain the exact
inspectable implementation.