Skip to main content

How to Manually Create a Photogrammetry Data Model

Updated

Summary: This article explains how to manually create a photogrammetry Data model in SDX-Cloud using uploaded drone, viDoc or SDX-PocketScan images. It walks through each step of the process, from upload to configuration and monitoring.

When to use this guide Use the manual workflow when you don’t have an automated integration (for example DJI Docks or Scin App), or when you want to process images from a third-party source by uploading them directly into a project.

Introduction

Photogrammetry Data models in SDX-Cloud can be generated automatically using integrations like DJI Docks, the Scin App with viDocs, or manually via image uploads.

This guide focuses on the manual workflow and is ideal when automated workflows are not available or when third-party image sources are used.

Prerequisites

You only need an active Project in SDX-Cloud to create a photogrammetry Data model. Learn how to create a project in SDX-Cloud

Processing eligibility (important) Photogrammetry processing is currently only supported for image datasets captured using DJI drones, viDocs or SDX-PocketScan. If you have a different dataset source, please contact support to request processing eligibility.

Step-by-Step Instructions

Explore the interactive demo above to see how to initiate the manual import process and configure file settings. You can also follow the written guide below.

1\. Upload the Dataset

  • Navigate to the Files section of your Project.
  • Upload your dataset as either:
  • A .zip file containing all images, or
  • Plain image files into a folder.
  • Ensure all images intended for a single Data model are grouped in the same folder.
Tip: Keep your upload structure simple Use one folder per Data model and name folders clearly (for example by date or site). This makes it easier to select the correct dataset during import and avoids mixing multiple captures into one Data model.

2\. Import a New Data model

  • Go to the Data Import screen in your project.
  • Click the “+” button at the top right.
  • Select “Image Data” as the file type.

3\. Select the Dataset

  • Choose the uploaded folder or .zip file.
  • If you select a folder, a cost estimate is displayed based on image count and processing settings.
  • If you select a .zip, cost estimates are not shown up front because the archive must be extracted first.
Tip: Prefer folders when you want cost transparency Folder-based imports show estimates immediately, which makes it easier to adjust settings before you start processing.

4\. Configure Processing Settings

  • Choose the photogrammetry provider (for example, Processing by Sodex).
  • Adjust point cloud precision and density.
  • These settings affect credit usage.
  • Credit estimates appear only for folder-based imports.
Tip: Start with lower settings, then increase if needed Higher density/precision can improve detail, but it also increases credits and processing time. If you’re unsure, start with a conservative setting and iterate after reviewing results.

5\. Use PPK Processing

If your dataset includes GNSS observation files, you can enable Post-Processed Kinematic (PPK) during Data model import to improve the accuracy of image geotags.

  • The PPK step is available only when the provider is set to Processing by Sodex.
  • PPK updates the images’ EXIF GPS coordinates before photogrammetry processing starts.

For requirements and the full workflow, see: Using PPK Processing for Photogrammetry Data model Import

Tip: Use PPK to improve alignment on challenging sites PPK can help when GNSS during capture was unstable (for example due to obstructions), because it refines image geotags before processing begins.

6\. Select Ground Control Points

You can choose one of the following:

Tip: Use Project GCPs when you want consistency Project Ground Control Points are useful when you process repeated Data models over the same site and want consistent georeferencing across datasets.

7\. Set Data model Metadata

  • Add a Data model name and an optional description.
  • Configure:
  • Tags
  • Point Insertion Option
  • Creation Date
  • Selecting “Get creation date from images” uses the earliest image timestamp during processing.
Data volume impact All Data models—both the images and the generated 3D point cloud data—count toward your processed data volume.

8\. Import the Data model

  • Click “Import Data Model”.
  • If all inputs are valid, the Data model is created and listed in the table.
  • Use the refresh button above the table to update the processing status.

9\. Monitor Processing

  • Processing duration depends on:
  • Dataset size
  • Chosen processing settings
  • Processing may range from a few minutes to several hours.
  • Configure notifications in Project Settings to receive alerts when Data model processing finishes or fails in the selected project.
Important: Stuck processing If processing has not completed after 24 hours, submit a support request so we can investigate.

Common Issues or Tips

  • Use folders over .zip files for faster and more transparent processing estimates.
  • Confirm image grouping before import (one folder = one Data model).
  • Data model creation timestamps may update after processing when image dates are available.
  • Confirm your credit balance is sufficient for the selected processing settings.

Related Articles

\\Available translations\\: en-us

Related articles

Comments

0 comments

Please sign in to leave a comment.

Powered by Zendesk