A measure of compactness for 3D shapes (Q1591974)

The classical measure of compactness of a 3D solid is defined by the ratio \(A^3/V^2\), where \(A\) is the enclosing surface area and \(V\) is volume of the solid. It is dimensionless, invariant under geometric transformations, minimized by the sphere, but sensitive to noise (small perturbations of the surface). The author defines a measure of discrete compactness \(C_D\) for a solid composed of \(n\) voxels by the formula \(C_D=(A_c-A_{c_{\min}})/(A_{c_{\max}}-A{c_{\max}})\), where \(A_c\) is the sum of areas of the contact surfaces which are common to two voxels, \(A_{c_{\min}}=a(n-1)\) and \(A{c_{\max}}=3(n-n^{2/3})\) are the minimum and, respectively, maximum contact surface area for a solid composed of \(n\) voxels (\(a\) is the area of a face of the voxel). Properties of the measure of discrete compactness are studied and some applications to digital image processing are presented.