Material characterisation is one of the most important aspects of accurate numerical modelling; correct material properties must be obtained for the correct behaviour to be observed. Traditionally permeability is measured by applying a constant/falling head test to a material sample, where such tests may involve many samples at varying pressure gradients. However current X-Ray micro-tomography techniques allow us to avoid physical lab tests by providing the ability to reproduce a voxelised representation of the internal structure of a porous medium. The Lattice Boltzmann Method may then be used to model a pressure induced flow field within the sample so that permeability may be numerically approximated. Typically this process is carried out after a thresholding procedure has been applied to the voxelised geometry to split it into definite solid and void spaces, at the expense of accurate representation of the geometry. In an attempt to better represent the porous medium the Immersed Moving Boundary technique was applied in such a way that it partially applies the bounce back boundary condition so that the strength of this application scales with the porosity of a given lattice node. This allows us to consider directly raw voxel values, avoiding the need for any thresholding procedure. To validate this hypothesis two test cases were explored in 2D; flow past a periodic array of cylinders by use of a unit cell model, and flow through a simple heterogeneous porous medium. Results were compared with analytical expressions where available, and published expressions for permeability evaluation of porous media. Results were found to be in good agreement with the available expressions.