To attain the corresponding discrete convolution equation, we represent the scene $\Omega $ as a vector by stacking the columns of the matrix: Display Formula
$f=vec(\Omega )=[f(\phi 1,\theta 1)\u2026f(\phi 1,\theta A),f(\phi 2,\theta 1)\u2026f(\phi 2,\theta A),\cdots ,f(\phi R,\theta 1)\u2026f(\phi R,\theta A)]T,$(6)
where vec maps $\Omega R\xd7A$ to $\Omega RA\xd71$ by stacking the rows of a matrix in a vector, and $[\xb7]T$ is used for the matrix-vector transposition. Likewise, Display Formula$g=[g(\phi 1,\theta 1)\u2026g(\phi 1,\theta A),g(\phi 2,\theta 1)\u2026g(\phi 2,\theta A),\cdots ,g(\phi R,\theta 1)\u2026g(\phi R,\theta A)]T$(7)
and Display Formula$n=[n(\phi 1,\theta 1)\u2026n(\phi 1,\theta A),n(\phi 2,\theta 1)\u2026n(\phi 2,\theta A),\cdots ,n(\phi R,\theta 1)\u2026n(\phi R,\theta A)]T.$(8)