Photogrammetry |
The generation and interpretation of photographic images and patterns to generate 3D depth data
Photogrammetry |
Photogrammetry uses a series of images or recordings to create depth data these images are then compared against each other and used to project and build a 3D mesh, the more reference images and angles inputted the more accurate the mesh generated.
Photogrammetry is a post-processing method of 3D generation relying on software interpretation to find common features from a series of photo references, this requires the creator to capture potentially hundreds if not thousands of images in an orbital nature around and or above the target ensuring any areas of possible occlusion/line of sight is retained in order to generate a ‘perfect’ output.
Points of commonality are used to match images against each other and map out the orientation of the images and project them against each other, depth is obtained based on common features, these fixed points dictate scale and orientation, these photos must be of the same aspect ratio and under ideal lighting and set conditions as the texture is generated and mapped from the photographic data and any movement or changes to capture target can result in capture process failure.
The biggest advantage photogrammetry has over other Depth capture options is the level of texture detail that is retained as the photographic data is mapped onto the mesh and as almost any camera can be used as long as all images used for training are of the same scale and quality can be a cheaper more accessible method to replicate 3D models.
Examples of Photogrammetry Software:
Autodesk Revit Pro, Meshroom, RealityCapture and RealityScan (Mobile), Polycam (Mobile) are a few examples please bare in mind some of these require subscription plans/purchases.