Thickness measurements of objects, especially transparent and semi-transparent objects, are essential for their characterization and identification. However, in the case of occluded objects, the optical thickness determination becomes difficult, and an indirect way must be devised. Thermal loading of the objects changes their opto-thermal properties, which will be reflected as a change in their optical thickness. The key to quantifying such occluded objects lies in collecting these opto-thermal signatures. This could be achieved by imaging the changes occurring to a probe wavefront passing through the object while it is being thermally loaded. Digital holographic interferometry is an ideal tool for observing phase changes, as it can be used to compare wavefronts recorded at different instances of time. Lens-less Fourier transform digital holographic imaging provides the phase information from a single Fourier transform of the recorded hologram and can be used to quantify occluded phase objects. Here we describe a technique for the measurement of change in optical thickness of thermally loaded occluded phase samples using lens-less Fourier transform digital holography and machine learning. The advantage of the proposed technique is that it is a single shot, lens-less imaging modality for quasi-real-time quantification of phase samples behind thin occlusions.