Table 4.

The 3D convolution (averaging) filters applied to 3D hyperspectral images (bands 0, 1, and 2) [41].

Data Size (${\mathit{N}}_{\mathbf{rows}}\times {\mathit{N}}_{\mathbf{cols}}\times {\mathit{N}}_{\mathbf{bands}}$)	($3\times 3\times 3$) Averaging Classical/ Noise-Free	($3\times 3\times 3$) Averaging Noisy (${10}^6$ Shots)	($5\times 5\times 5$) Averaging Classical/ Noise-Free	($5\times 5\times 5$) Averaging Noisy (${10}^6$ Shots)
($8\times 8\times 8$)
($16\times 16\times 16$)
($32\times 32\times 32$)
($64\times 64\times 64$)
($128\times 128\times 128$)