Unit 1.1-SfM/GNSS: Structure from Motion with UAV, GNSS and ground control points (GCPs)

Sean Gallen and students Maggie Windingstad, Nathan Weaver, and Emily Hofer; Colorado State University

Initial Publication Date: August 10, 2025 | File/Data Set Update: September 5, 2025 (see revision history: 4 events)

First Publication: August 10, 2025
File/Data Set Update: August 13, 2025 -- Updated file: 'MetaShape Guide for SfM with GNSS-georeferenced GCPs' Updated to new tool menu configuration.
File/Data Set Update: September 5, 2025 -- Updated file: 'MetaShape Guide for SfM with GNSS-georeferenced GCPs UAV survey'
File/Data Set Update: September 5, 2025 -- Revised version of MetaShape guide and additional suggestions in the Description.

Summary

In this unit, students learn how to design and conduct a Structure from Motion (SfM) survey with ground control points (GCPs). Photos are collected with an uncrewed aerial vehicle (UAV). The survey is geoferenced using real-time kinematic global navigation satellite system (RTK GNSS) instruments. The photos are processed using Agisoft Metashape software to build a variety of 3D models (point clouds, etc.), which students can analyze to address a geologic research question. A wide variety of societally valuable potential science topics are suggested.

SfM is a photogrammetric method that constructs 3D models using overlapping photos. SfM applications are widespread and commonly used in numerous subdisciplines in Earth Science. RTK GNSS techniques allow rapid collection of high-accuracy and high-precision (±2 cm) position information. Combining SfM and RTK GNSS is commonly used to locate SfM models precisely and for temporal surveys in the exact location for surface change detection.

Show more information on GPS versus GNSS

Note: Although the term GPS (Global Positioning System) is more commonly used in everyday language, it officially refers only to the USA's constellation of satellites. GNSS (Global Navigation Satellite System) is a universal term that refers to all satellite navigation systems including those from the USA (GPS), Russia (GLONASS), European Union (Galileo), China (BeiDou), and others. In this unit, we use the term GNSS to refer generically to the use of one or more satellite constellations to determine position.

Geospatial, Geodesy, GIS/Mapping/Field Techniques, Geomorphology | College Upper (15-16) | Designed for In-Person

Expand for more detail and links to related resources

Share your modifications and improvements to this activity through the Community Contribution Tool »

Learning Goals

By the end of this unit, students will be able to:

Design and conduct a combined SfM-GNSS survey considering site-specific characteristics and science objectives, including:
- List the equipment needed for an SfM UAV survey with GCPs.
- List and carry out the steps needed to conduct a UAV-platform SfM survey with GCPs (survey workflow).
- Use best practices in designing layout, distributing, and surveying GCP targets.
- Make supporting field notes and observations required for survey analysis and interpretation.
Process SfM photos with GCPs to generate 3D models, including point clouds, digital elevation models, and orthomosaics, that can be used to address a geoscience research question (model workflow).

Context for Use

This unit is designed for upper-level undergraduate and graduate students in the geosciences, civil engineering, or related fields. It is designed for small-to-medium-sized classes (~5-20 students) with an instructor and, ideally, a teaching assistant for larger classes. It is appropriate for academic year courses with field components or a component of a summer field camp.

This is a more advanced unit that links components of GETSI's field modules High Precision Positioning with Static and Kinematic GPS and Analyzing High Resolution Topography with TLS and SfM. This unit assumes completion of Unit 1: GPS/GNSS Fundamentals andUnit 2: Kinematic GPS/GNSS Methods in the GPS module and builds on Unit 1-SfM: Introduction to SfM in this module. Students should be confident working with kinematic GNSS instrumentation.

The unit requires ~20 minutes for the initial lecture, ~40 minutes for in-class survey pre-planning and class discussion, and ~2-3 hours for field survey planning and implementation. Post-survey data processing should take ~3 hours with teaching materials provided. If implemented during the academic year, it is recommended that this activity be divided into three sessions: introduction and planning (in-class and field), survey and data collection (in the field), and data processing and analysis (in a computer lab or with student laptops).

Description and Teaching Materials

This unit can be adapted to a wide variety of field sites or prepared data sets. Review Units 2-5 in this module for ideas of sites that will work best for your learning goals. Then use the survey design guide here to carry out a combined SfM-GNSS survey using a UAV collection platform.

Description

This unit has three stages, reflective of the survey stages: (1) introduction and pre-survey planning, (2) conducting the survey, and (3) processing the data.

Introduction and pre-survey planning

The pre-class reading, introductory lecture, and in-class exercise are preparation for planning the field survey. The reading material guides students through the thought process of how to collect data effectively (and considers how it will be processed in the lab). The reading emphasizes how survey design can affect final results and requires students to consider variables related to logistics, safety, and best practices in survey design. Reading materials should be summarized in class, given the lecture materials provided. This should be followed by a class discussion of survey design, which includes collaboratively developing a field survey plan using maps and Google Earth of the study site. The provided PowerPoint presentation, student reading, pre-trip planning exercise, and Google Earth file can be used to practice the process at the example field site before doing pre-planing for the actual field site. Or the instructor could choose to adapt the files to feature the actual course field site from the beginning.

Conducting the survey

The survey should be conducted during one or two lab periods (if during the regular semester). First, the instructor and/or student must acquire and check equipment for the survey (e.g., UAV, RTK GPS/GNSS units, GCP targets, batteries, first-aid kit, water, food, etc.). At the field site, the students and instructor should assess the pre-survey plan, revise as needed, and decide on a staging area and bring equipment there. Students should divide into groups to disperse the GCP targets according to the survey plan and measure their precise location with the RTK GPS/GNSS units while taking notes. The UAV survey occurs in tandem with or following measurement of the GCPs. After the survey, students should collect and inventory all field equipment before packing the vehicles and departing the field area. Instructors can request support for GPS/GNSS instrument loans from EarthScope Consortium, which runs the NSF's geodetic facility.

Processing the data

In a computer lab, this portion of the unit should take place during one or two lab periods. Computers must have Agisoft Metashape, a licensed software, if using the guide provided with these materials. Before class, the instructor should compile the photos from the UAV survey and the RTK GPS/GNSS data and provide this data to the students. The instructor should give a brief <5 min description of the goals and expected outcomes of the activity. The students can then work individually or in groups to work through the workflow provided. The instructor should circulate around the room to help students troubleshoot as needed.

1) Pre-survey Planning

The instructor presents students with a field area and asks them how they would best design an SfM survey with ground control points (GCPs), where the GCPs are surveyed with kinematic GNSS. What are the important considerations? Where would they place the base station and why? What are the best positions and materials for the GCP targets*, given survey needs, logistics, and safety? Students should first work individually, then in small groups to discuss the survey designs, before convening as a class to discuss and arrive at a consensus on the survey design plans.

*GCPs are typically identified with targets – physical markers that can be identified in the SfM photos – that are surveyed precisely with high-precision GPS/GNSS. Targets can be as simple as a clearly identifiable ground feature, like a lone boulder in a grassy field, but are more commonly temporary features, like flags, flat-lying fabric markers, or spray-painted X's.

Show more guidance for Pre-survey Planning

It is recommended that the instructor print images of the study area (with a scale bar) from Google Earth or a GIS platform and distribute them to students to work with in class. With Google Earth on the screen and the provided imagery handouts, lead a discussion on the science objectives. It is critical to consider what can be done with this data, what you want to do with it, and what resolution of the data is needed to achieve the objectives. Once the objectives are settled, it is time to pre-plan the survey.

Important questions to address include:

What are the necessary safety protocols and safety equipment? What are potential hazards and mitigration? (e.g., deep water, wildlife, lightening, rock falls)
What scientific equipment do we need? (having students develop a checklist can be very effective)
How will we dress and prepare (e.g., weather, water, and food)?
Where will we park, set up, and test the equipment?
Do we have access to the field site (e.g., avoid private property)?
What is the scale of the survey, and can we tentatively define the area of interest?
What resolution is required for the SfM model to meet the study objectives?
Are there obvious obstructions (i.e., power lines, buildings, etc.) that might interfere with the planned survey?
Where should the RTK GNSS base station be set up? (some place not to difficulat to get to, that it has clear sky view and line of sight to GCP targets; also the base station should not fall over, get damaged, or get stolen)
What is a good design for the GCP target locations?
How long will the anticipated survey take?
How will the work tasks and equipment responsibility be divided?

Either have the students work in groups or work as a class to address the above questions and put points on the map for the planned survey. If done in groups, each group should present their design plan, and the class should work together to find agreement on the best survey strategy. Remind the students that having GCPs in a grid-like pattern that spans the perimeter and center of the survey area will produce the best results.

2) Conducting the Survey

×

SfM with GCP survey workflow
Provenance: Sean Gallen (Colorado State University)
Reuse: This item is offered under a Creative Commons Attribution-NonCommercial-ShareAlike license http://creativecommons.org/licenses/by-nc-sa/3.0/ You may reuse this item for non-commercial purposes as long as you provide attribution and offer any derivative works under a similar license.

Students will work in groups to implement the survey design for an SfM digital survey model with GCPs. Students are responsible for all aspects, from testing, packing, and setting up equipment to surveying and data collection (except for the UAV flight, which will be conducted by the instructor or technician with an appropriate FAA UAV license). It is best for students to divide tasks into groups and work as a team to efficiently conduct the survey (e.g., placing GCP targets, setting up and testing the base station and rover, and assessing the SfM flight path for on-site conditions). Students must collect well-organized and complete field notes, including appropriate field sketches, metadata, and logs, which will be submitted and assessed.

Show more guidance for Conducting the Survey

Before leaving for the field site, ensure that you check the equipment against the equipment list to confirm you have all the necessary items. You can also task students with preparing the list and/or checking that everything is packed and ready. Check that the batteries are charged and that the firmware has been updated the night before. Ensure students are prepared for the field and verify that you have packed the necessary safety equipment (e.g., first aid).

At the field site, introduce the students to the field area and review the science objectives before taking any geodetic equipment out of the vehicle. If the survey area isn't far from the parking area and isn't too large, it is useful to conduct a walk-through before removing the equipment from the vehicles. If the survey site is far away, a walk-through is still extremely useful and recommended; however, it is best to move the equipment before starting the walk-through. During the walk-through, have the students take notes relevant to their science objectives and have them review the survey pre-plan. If any changes are needed to the pre-plan, make them now before deploying the equipment.

After the introduction and walk-through, verify that the equipment matches the equipment list to ensure everything necessary is present. Introduce students to each instrument (RTK GNSS, UAV, tablets) and associated equipment (tripod, monopod, targets). Be sure to include instructions on how to care for the instruments to prevent damage during fieldwork. Show them how to use the equipment and supplies they haven't used before. After testing the equipment in the staging area, have the students finalize the locations where they want to set up the RTK GNSS base station and GCP targets, taking into account considerations such as the stability of the site for the base station tripod and the line of sight with the GCP targets. Deploy the students to efficiently set up the base station and targets. Students deploying the targets need to ensure that the center of the target has something hard underneath it. If not on bedrock, a large pebble or small cobble can be placed under the target center. This ensures stability when surveying points with the rover. If planning to resurvey the same site, students might want to mark GCP target locations in some way, if allowable at your field site. For example, with spray paint, a survey pin, or driven rebar rod. It is also useful to number the GCP target locations. Most commercial GCP targets already have a sequential number, making this process easy. Be sure to note the GCP and target number, as well as their locations, on a map and in field notes for later reference.

The individual or group setting up the base station should maintain communication with the roving station (e.g., via a walkie-talkie or cell phone) and turn on the base station so that it starts collecting data. The group with the rover can then turn it on to make sure the base station and rover are communicating.

Once the targets are out, the base station is up, and base-rover communications are established, the RTK GNSS survey of the GCP targets can begin. This is best done in pairs or groups, where one student works with the rover on a monopole or bipod, and another works with the associated tablet and takes field notes. When collecting roving data, the student should record notes in the controller (typically a phone or tablet) related to the GCP number and also take field notes with the same information. Additionally, they can annotate information on a map if needed. Collect RTK GNSS locations for all GCP targets and return to the staging area.

While students are placing and surveying GCPs, another group can plan the UAV flight based on the area of interest, desired model resolution, and other site-specific considerations (e.g., weather, trees, buildings, powerlines) . Most UAVs have software to automate this process, making it easy to design flight paths. The pilot must be aware, however, of local obstructions and weather conditions before executing a flight plan. Once students come back from the surveying GCP targets are in the staging area, the UAV pilot will execute the flight and collect the photos. It is sometimes useful to take notes on changing weather and light conditions, making sure to precisely note the time of any relevant changes, as they can affect photo quality and thus the resultant SfM model.

Once the UAV survey is completed, students can be deployed to retrieve the targets and then turn off and dismantle the base station. Before loading the equipment into the vehicle, compare it against the equipment list to ensure nothing is left in the field.

3) Data Processing and Assessment

×

SfM-GCP data processing workflow
Provenance: Sean Gallen (Colorado State University)
Reuse: This item is offered under a Creative Commons Attribution-NonCommercial-ShareAlike license http://creativecommons.org/licenses/by-nc-sa/3.0/ You may reuse this item for non-commercial purposes as long as you provide attribution and offer any derivative works under a similar license.

After students have implemented and completed their survey, they should hand in their field notebooks for assessment and feedback. The students will process the kinematic GPS data and the SfM data with GCPs using the guide provided, which walks students through the SfM with GCP processing using Agisoft Metashape. Students should prepare a written description and analysis of the lessons learned that summarizes their experience and knowledge of kinematic GNSS and SfM survey design and implementation with GCPs based on the Summative Student Exercise. It is recommended that students work individually to analyze and interpret their data based on the specific objective of the survey. This portion of the activity is class- and topic-specific, so the instructor is advised to adapt the activity as appropriate.

Show more guidance for Data Processing

The instructor should ensure that students or groups take the lead on a given dataset (e.g., GCPs from the RTK GNSS or photos from the UAV) and store and upload these files to a location accessible to the entire class. The instructions for this activity focus on generating an SfM model using Agisoft Metashape; however, other packages, such as OpenDroneMap and Pix4D, can also generate SfM models with GCPs. This part of the unit is designed to be delivered to a computer room with SfM software (alternatively, free 30-day trial licenses can be used on student laptops). Students can work individually, in pairs, or in groups, but we recommend working in pairs to facilitate troubleshooting and maintain an active learning environment conducive to peer-to-peer learning. This part of the unit is best delivered during a lab session or at least two classroom sessions.

At the beginning of class, review the science objectives and the datasets that have been collected. Discuss the overarching goals. Make sure students have access to the data. Cover the basic workflow for generating an SfM model with GCPs. There is no need to go into too much detail at this point; simply outline the workflow steps. The students can then follow the guide provided. As the students work through the manual, circulate around the room and help troubleshoot and answer questions. If common issues or questions arise, address them with the attention of the entire class.

Teaching Materials

Pre-survey planning - These planning documents can be used to practice the process at the example field site before doing pre-planing for the actual field site. Or the instructor could choose to adapt the files to feature the actual course field site from the beginning.
- Introduction to SfM with UAV GNSS and GCP surveying presentation (PowerPoint 2007 (.pptx) 13.4MB Sep5 25)
- Pre-class reading on SfM-GNSS surveying with GCPs (Microsoft Word 2007 (.docx) 5.8MB Aug9 25)
- Student exercise: SfM-GNSS pre-trip planning activity (Microsoft Word 2007 (.docx) 4.3MB Aug9 25)
  - Example Survey Area.kmz (KMZ File 1kB Sep5 25)
Conducting the survey, Data processing, and Assessment
- SfM survey using GNSS-GCP-UAV summative student exercise (Microsoft Word 2007 (.docx) 240kB Aug10 25) - This file provides a framework of a student exercise for a UAV SfM survey where the ground control points are georeferenced with GNSS. Instructors using this unit should plan to fill in details about their particular field setting and science question/s.
- MetaShape Guide for SfM with GNSS-georeferenced GCPs (UAV survey) (Microsoft Word 2007 (.docx) 33.4MB Sep5 25)
- Unit 1.1 rubric (Microsoft Word 2007 (.docx) 234kB Aug10 25) - This rubric has suggested grading criteria for the student planning, field, and summative exercises of the unit.

Teaching Notes and Tips

Handling group sizes: With smaller groups, students will be introduced to the equipment and gain hands-on experience with each aspect of the survey. With larger groups, it is best to break them into smaller groups and introduce the equipment by rotating instructors. Furthermore, for larger classes, groups can be tasked with different components of the survey (e.g., placing GCP targets, setting up the RTK GNSS, testing the UAV, RTK GNSS survey of GCP targets, setting up and conducting the UAV survey) to ensure each group is actively involved in some aspect of the survey. In such cases, it is recommended that students from each group give presentations at the end of the day or the following class periods, explaining to the other students what their group has done. Equipment introduction can be done in the classroom or in the field. Aspects of survey pre-planning are best done in the classroom with the aid of Google Earth, but should be revisited and likely modified upon arrival at the field site depending on site conditions (e.g., obstructions, weather).
Processing time: SfM post-collection processing takes several hours to days to complete. SfM post-processing is best done in a computer lab during another class or lab period. Students can work independently, in pairs, or in groups to generate 3D models. For larger groups, we recommend that students work in pairs, as computer skills are likely to vary. Prior to class, instructors should work to ensure quality control of the data collected and ensure that all data is easily accessible to all students before the start of the class or lab period, particularly for larger classes where various groups collect different data.

Share your modifications and improvements to this activity through the Community Contribution Tool »

Assessment

Formative Assessment:

Much of the formative assessment can be done through observations of and discussions with students individually, in pairs, or periodically in the whole group. This could include:

Introductory lecture: Class discussion, think-pair-share, and/or in-class planning activity
In the field: probing questions, guided setup, and survey design

Summative Assessment:

A component of the summative assessment is the pre-trip planning activity exercise, in which students demonstrate their ability to design a plausible SfM/GNSS survey with GCP. A summative assessment for a general report that describes the survey workflow and the basic steps to generate an SfM model with GCPs is provided in the Summative Student Exercise. For a more thorough summative assessment, select a target science topic from the Analyzing High Resolution Topography Units 2-5 and adapt it to address a relevant field site for your course, or choose a prepared data set to have students work with.

Unit 1.1 rubric (Microsoft Word 2007 (.docx) 234kB Aug10 25) - This rubric has suggested grading criteria for the student planning, field, and summative exercises of the unit.