In our previous paper, we predicted neutron skin r skin and proton, neutron, matter radii, rp, rn, rm for 62,64 Ca after determining the neutron dripline, using the Gogny-D1S Hartree-Fock-Bogoliubov (GHFB) with and without the angular momentum projection (AMP). Using the chiral g-matrix folding model, we predicted reaction cross section σR for 62,64 Ca scattering on a 12 C target at 280 MeV/nucleon, since Tanaka el al. measured interaction cross sections σI(≈ σR) for 42−51 Ca in RIKEN. After our prediction, they determine rm(RIKEN), r skin (RIKEN), rn(RIKEN). In this paper, we reanalyses the σI, since they assumed the Wood-Saxon densities for 42−51 Ca. The σR calculated with the folding model with GHFB and GHFB+AMP densities almost reproduce the σI. We then scale proton and neutron densities so that rp and rn may agree with the central values of rp(exp) and rn(RIKEN), respectively. The σR calculated with the scaled densities do not reproduce the central values of σI perfectly. We then determine the rm that agree with the central values of σI, using the chiral g-matrix folding model. The fitted rm do not reproduce the central values of rm(RIKEN) perfectly, but are in one σ level. Finally, we show the r skin , rn determined from the fitted rm are close to the original ones except for r 48 skin . The fitted r 48 skin is 0.105 fm, while the central value of r 48 m (RIKEN) is 0.146 fm. When we fit rm to the upper bound of σI, the fitted r 48 skin is 0.164 fm and near the central vale 0.17 fm of the high-resolution E1 polarizability experiment.