I've been experimenting the last few days with Gleb Alexandrov's free-software techniques for photogrammetry capture. I've added a couple of steps of my own however.
I'm shooting on a Canon T3i DSLR, and using Magic Lantern firmware which allows me to shoot Dual ISO HDR images. I run these through the CR2HDR processor, which allows me to take my captures into Adobe Camera RAW and get a high quality image which minimises lost shadow details. I then save those images out to a folder for VisualSFM to reconstruct my point-cloud, which is then sent to MeshLab to generate the geo. I generate 2 sets of geo, my low-poly using a reconstruction depth of 7 or 8, and my high-poly at a depth of 12 or 15. From my high-poly I generate the textures. I export both meshes out as OBJ, and take the lo-poly into Blender quickly to generate a UV unwrap for it. Finally, I take these into Substance Designer 6 to do my baking and PBR texture set creation, with some Bitmap2Material for things like roughness.