Best photogrammetry software

When tolerance benchmarks are set at 1/500 of a millimeter, you might want to reach for metrology-grade photogrammetry gear.

For the purists, the term is used strictly for the process of taking reliable measurements from photographs. Photo – light, gramma – writing, and metron – to measure. Measurement using light. Its use is more scientific – more industrial, if you will. You’ll find it in applications like inspection, deformation analysis, and other high-accuracy measurement use cases.

For many in visual arts and heritage preservation, for example, its definition is simple: turning pictures into 3D models. Historians use it to digitize archaeological specimens in full color, CGI artists to quickly build assets for animation or game development, and – some hobbyists – to 3D-print models for the sole purpose of showing off to their mildly interested dinner guests.

We’ll call it a technique for reconstructing 3D data from 2D images. This definition should cover both photogrammetry for metrology, and what we’ll refer to as photogrammetry for CGI.

Barack Obama sitting in front of a custom-built photogrammetry setup consisting of 50 LED lights and 14 cameras. Taking multiple shots of the president’s face from different angles in a matter of seconds allowed for obtaining high-quality texture data and later combining it with 3D scans made with two Artec Eva scanners, resulting in the first-ever full-color 3D portrait of a head of state. Photo courtesy of the White House.

3DF Zephyr

3DF Zephyr comes in three versions. At the top of the range, there is the full version, which you can pay for as a monthly subscription, or one time for a perpetual license. There is also 3DF Zephyr Lite, which occupies the middle tier, and 3DF Zephyr Free.

The free version gives you the ability to construct 3D models but confines the number of photos you can use to just 50. This limits the sizes of objects you can work with. You can also only use one NVIDIA GPU to enhance processing, and the options for export are restricted to just meshes with JPEG texture and uploads to YouTube. 3DF Zephyr Free also provides limited editing tools to enable you to do just the basics – selections, cutting, and copying elements, and the like.

While these restrictions may seem inhibiting, it is still a great place to start if you just want to experiment and get a feel for the software. You can later upgrade to higher tiers without these limits and take advantage of a greater selection of professional editing tools and filters, wider export options with point clouds, textured meshes, positions and distances, and more, and features like support for laser scanner file formats, the ability to generate orthophotos, as well as tools for surveying, GIS, and CAD.

Download link: https://www.3dflow.net/3df-zephyr-free

Meshroom

This is an open-source solution that is built on AliceVision, a framework of computer vision algorithms with 3D reconstruction and camera tracking. It has an easy to use drag-and-drop editor that makes working with the software intuitive. It uses a node-based workflow, where each node represents a set of actions to be performed on the way to generating a 3D model.

The software gives you the ability to tweak the workflow – or pipeline – or simply work with the default, one-click workflow. This flexibility makes Meshroom both beginner-friendly and beneficial to more advanced users. You can also save your custom pipelines as templates so you don’t have to retrace your steps in subsequent projects.

Meshroom can support over 1,000 images and is compatible with other 3D modeling software. This gives users more flexibility around the workflow. You can use it to, for example, add texture to a mesh after editing it in other 3D modeling applications.

A range of plugins allows for cross-compatibility with software such as Houdini, Blender, Maya, ZBrush, and more for more advanced workflows for game development, photo modeling, and camera tracking. Users also benefit from libraries with additional algorithms for 3D reconstruction, target identification, and more.

VisualSFM

The SFM in this solution’s title stands for Structure from Motion. It is a technique for inferring geometrical information by moving around an object. To oversimplify the premise behind it, one could say it works by quantifying motion parallax. As an observer changes position, objects that are closer to them seem to shift more than those that are further away. If you do the math from these relative differences, you can determine the 3D structure of what is being observed. Human beings use the technique too.

VisualSFM gives a graphical user interface to these calculations. It is a no-frills solution built by Google software engineer Changchang Wu, using SFM calculations to feature detection, feature matching, and bundle adjustment to enable users to generate 3D models from 2D images.

The process itself is completed in four steps, which can also be run from the command line if you prefer that over the GUI. Users import images in VisualSFM, run feature detection and image matching, sparse reconstruction follows next, and finally, dense reconstruction.

VisualSFM is free for personal, non-profit, or academic use.

Download link: http://ccwu.me/vsfm/index.html

The pool of options if you are looking for a photogrammetry solution is wide, and growing. The solutions we have listed here are a great place to start your search for a solution that fits you. In terms of functionality, the options we have covered are fairly representative of what competing products on the market will typically offer. The choice you finally settle on will, of course, be determined by your needs, level of expertise, available resources, and personal preferences.