pub fn plan_decimated_blur(std_deviation: f32) -> (usize, [f32; 13], u8)Expand description
Compute the blur execution plan based on standard deviation.
Returns (n_decimations, kernel, kernel_size):
n_decimations: Number of 2× downsampling steps to perform (per axis)kernel: Pre-computed Gaussian kernel weights (fixed-size array)kernel_size: Actual length of the kernel (rest is zero-padded)