Skip to main content

v1.1

Minor Versionm

by Eli Jared Fastow

Introduction

Photogrammetry consists of processing a large number of photographs of an object from a variety of angles into a 3D mesh. This guide provides a basic introduction to the use of Meshroom, a free and robust photogrammetry software.

    • Photogrammetry is far less accurate and precise than any of the other scanners available at Terrapin Works

    • If another scanner is available and the object fits within its appropriate use-case, use a designated 3D scanner.

    • Objects not reasonable to scan with designated 3D scanners, like buildings, trees, or other large objects may fit within the use case for photogrammetry

    • Photogrammetry also works for objects with textures not amenable to the other 3D scanning technologies, like plants.

  1. Successfully constructing a 3D mesh of an object from photogrammetry requires capturing dozens of pictures from all relevant angles. The term all relevant angles typically means a 360 view of the object, but for a target where all the important features are on one side - like a human face - only focus on that area.
    • Successfully constructing a 3D mesh of an object from photogrammetry requires capturing dozens of pictures from all relevant angles.

    • The term all relevant angles typically means a 360 view of the object, but for a target where all the important features are on one side - like a human face - only focus on that area.

    • The field of view for adjacent photographs should overlap by at least 50%

    • Make sure you have access to take a picture of the object from all relevant angles

    • Try to keep the lighting on the object as uniform as possible

    • Keep the background as featureless as possible

    • Try to reduce deep shadows on the object

  2. Take photographs across the range of relevant angles  on multiple planes
    • Take photographs across the range of relevant angles on multiple planes

    • Make sure each photo overlaps by about 50%

    • Take pictures on a maximum depth of field

    • Attempt to reduce photographic aberrations, including motion blur, lens flares, over/under exposure, and poor focusing

    • Save all photos as high resolution images. Compressed photos can create artifact features that Meshroom may mistake for real objects.

  3. Catalog of all images taken
    • Catalog of all images taken

    • Display of selected image

    • Display of generated mesh

    • Mesh settings

    • Processes used to convert images to mesh

  4. Save all of the pictures in a single folder, then drag-and-drop the folder into meshroom Click on any individual image in the image catalog to view it at full resolution Check all images are clear, focused, and centered
    • Save all of the pictures in a single folder, then drag-and-drop the folder into meshroom

    • Click on any individual image in the image catalog to view it at full resolution

    • Check all images are clear, focused, and centered

    • Click the start button to run the algorithm that converts images to mesh files. Note that this process may take a while and use lots of CPU resources.

    • As the algorithm runs, it will begin to build a mesh. Once the algorithm finishes running, press the load model button

    • Track progress with the processes window at the bottom. The orange bar indicates the step currently operating

    • After clicking the load model button, a mesh file is automatically exported. Click the open containing folder button to see it.

Finish Line

One other person completed this guide.

Eli Jared Fastow

Member since: 11/04/2019

3,602 Reputation

30 Guides authored

0 Comments

Add Comment