------------------------------
VIP Info:
	 Computes a differential invariant
------------------------------
For the Gaussian filtering Sigma is adapted to the voxel size in each direction
The result is a float Tivoli/GIS format

Usage: VipGeometry
        -i[nput] {image name}
        [-m[ode] {mc/gc/mlvv/glvv/pc1/pc2/la/nla  (default:mc)}]
mc = isophote mean curvature, gc=isophote Gaussian curvature
mlvv = isophote mean Lvv, glvv=isophote Gaussian Lvv
pc1 = isophote principal curvature 1, pc2=isophote principal curvature 2 (pc1>pc2)
la = Laplacian, nla = Laplacian (Normalized Laplacian is not implemented)
        [-s[igma] {float (mm)  (default:largest pixel size)}]
Apply a smoothing (scale=sigma*sigma) to compute differentials in scale-space
        [-g[aussian] {y/n (default:y)}]
        [-H[isto2D] {y/n (default:n)}]
        [-G[radient] {y/n (default:n)}]
        [-M[asked] {image name for Histo2D (default:"input")}]
        [-T[hreshold] {on masked image for Histo2D (default:"1")}]
        [-G[inf] {float   (min of geometry for Histo2D (default: -10))}]
        [-G[sup] {float   (max of geometry for Histo2D (default: 10))}]
        [-o[utput] {image name (default:"geometry")}]
        [-r[eadformat] {char: v or t (default:v)}]
        [-h[elp]