AT 309 Final Data Product GIS and Mapping Introduction

AT 309 Final Data Product GIS and Mapping

Introduction

In AT 309, multiple different labs were completed with different aircraft to gain experience working with UAS data, GIS software, and 3D mapping products. Some of the aircraft used includes the DJI Mavic 2, Skydio 2+, and the DJI Matrice 300. These labs focused on collecting aerial imagery and processing it into usable outputs such as Ortho mosaics, digital surface models, digital terrain models, and 3D points clouds. We went to various locations to allow different series of data to be collected, these location included the Student Garden at Purdue University, and the Turf farm research and diagnostic field at Purdue university. The goal of these labs was to understand how raw drone data can be turned into accurate maps and spatial data for real world applications.

Throughout the semester, we worked with software like Drone2Map, ArcGISPro, ArcGIS Earth, and Pix4D. Drone2Map and ArcGISPro were used to process imagery, manage coordinate systems and projections, and apply basic cartographic standards. The labs also highlighted how flight planning, overlap, and data quality affect the final outputs. Overall this report summarizes the workflows used across all labs, the skills developed, and the results of the GIS and 3D products created during AT309.

Processing insight

Overall steps we took to process data for the three data sets. To begin processing the UAS imagery, each mapping mission was first organized into separate data folders to ensure the datasets remained clear, readable, and easy to manage throughout processing. Once organized, the imagery was uploaded into Drone2Map, where the appropriate processing workflow was selected. For these projects, the focus was on generating standard 2D mapping products, so the Ortho mosaic, DSM and DTM outputs were enabled. Before starting the processing, basic project settings were reviewed, including image alignment options and output selections. After confirming the desired products, the processing was initiated and monitored as the software completed image alignment, dense matching, and final product generation. Once processing finished, Drone2Map produced a 2D map view displaying the completed Ortho mosaic and elevation models.

After the Drone2Map processing was complete, the final outputs were transferred directly into ArcGISPro for visualization analysis, and map creation. In ArcGISPro, multiple layouts were created to examine and compare the different data products. The DSM was used to visualize the full surface of the area, including terrain, buildings, vegetation, and other objects, making it useful for applications such as urban planning, emergency response, and infrastructure analysis. The DTM was used to represent the bare-earth surface by removing above ground features, which is especially important for terrain analysis, engineering, and construction-related work. To improve interpretation of elevation changes, shaded versions of both the DSM and DTM were created using hill shading, which simulates sunlight and shadow to give the terrain a more 3D appearance. In addition, an Ortho mosaic map was created, which combines overlapping aerial images into a single, geometrically corrected image that allows for accurate measurements and visual assessment of the area.

For each project Drone2Map also generated supporting processing information such as image count, processing time, cell size, pixel depth, elevation values, and coordinate projection details. Because the same general processing steps were applied across multiple mapping missions, it was possible to compare the quality of the final products between different flight patterns and conditions. This workflow demonstrated how consistent processing settings, combined with good flight planning, result in higher-quality spatial data that can be effectively used within a GIS environment

Data set 1 DJI Mavic 2 Red boundary

For this lab, our team conducted a mapping mission using the DJI Mavic 2 Pro to collect aerial imagery over the assigned study area at the Purdue Student Garden. The mission was flown using a parallel “lawn mower” grid pattern, which allowed for consistent image overlap and efficient coverage of the site. This lab also introduced a new flight application that does not require a paid subscription, providing an opportunity to compare its usability and performance to other mapping platforms used in previous missions, such as the Skydio s2. The primary goal of this mission was to gather high quality imagery for future processing and to gain hands on experience with mission planning, execution, and metadata collection for grid based mapping. By flying a single area using the Mavic 2 Pro, this lab emphasized understanding how flight parameters, overlap settings, and environmental conditions influence data quality while ensuring safe and compliant UAS operations.

Weather

Table 17

Table 17

Temperature

Visibility

4 miles

Precipitation Prob.

38%

Wind

7 MPH

Gusts

KP index

Sats

There are no rows in this table

⁠

Parameters for flight

Table 18

Table 18

Altitude

80 meters (200 feet)

Overlap

80%

Camera orientation

Nadir

There are no rows in this table

⁠

DSM

⁠

Figure 1 DSM Red boundary

Table 19

Table 19

Cell Size

X= 1.647cm Y= 1.647cm

Pixel Depth

32 bit

DSM Value

202.344 -246.642

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Review of DSM for Red Boundary DJI Mavic mission

Surface height variation is clearly visible, especially where structures and vegetation exist.

Patchy areas and speckling appear in some sections, indicating less reliable elevation estimates.

These inconsistencies likely result from reduced overlap and lack of cross-hatch flight paths.

Overall, the DSM is usable but less refined than a full grid mission.

Shaded DSM

⁠

A map of a red and green color

AI-generated content may be incorrect.

⁠

Figure 2 Shaded DSM Red Boundary

Table 20

Table 20

Cell Size

X= 1.647cm Y= 1.647cm

Pixel Depth

32 bit

DSM Value

202.344 -246.642

Shaded DSM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DSM review

The DSM clearly captures surface features such as buildings, tree lines, and vegetation height changes.

Flight liens and image centers are consistent, showing good coverage and overlap.

Higher elevation values appear along the western edge, likely caused by trees and uneven vegetation.

Some noise and rough texture near the boundary edges suggest weaker overlap and edge-of-mission artifacts, which is common in boundary-only flights.

DTM

⁠

Figure 3 DTM Red Boundary

Table 21

Table 21

Cell Size

X= 8.2337cm Y=8.2337cm

Pixel Depth

32 bit

DTM Value

202.419 – 222.546

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Review for DTM

The terrain surface is smoother than the DSM and highlights overall ground shape well.

Elevation transitions are more uniform across the site.

Some distortion near the edges suggests insufficient ground point classification

This directly reflects the limitations of collecting only boundary data rather than the full coverage.

Shaded DTM

⁠

Figure 4 Shaded DTM Red boundary

Table 22

Table 22

Cell Size

X= 8.2337cm Y=8.2337cm

Pixel Depth

32 bit

DTM Value

202.419 – 222.546

Shaded DTM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Review for shaded DTM

The DTM does a good job removing most above-ground features, leaving a smoother representation of terrain.

Gradual elevation change from west to east is visible, showing the natural slope of the area.

Minor rough patches remain near the western boundary, likely due to dense vegetation or limited ground visibility during collection.

These artifacts likely to relate to flying only a boundary instead of a full interior grid.

Ortho mosaic

⁠

Figure 5 True Ortho Red Boundary

Table 23

Table 23

Cell Size

X= 1.647cm Y = 1.647cm

Pixel Depth

32 bit

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Review for orthomosaic

The orthomosaic is visually clear with good color consistency and sharp detail.

Structures, roads, and field patterns align well with real-world features.

Minor stretching and distortion appear near the outer edges of the mosaic.

These edges issues are consistent with limited overlap and boundary-only mission planning.

Data Set 2 Skydio Red and Blue Boundary

This was my second data set that I processed through Drone2Map!

In this lab, our team completed two mapping mission using the Skydio platform at the Purdue Student Garden to practice planning, executing, and managing autonomous UAS mapping flights. We flew both a standard lawnmower grid mission and a cross-hatch mission with a perimeter scan to simulate real world mapping scenarios. The lab focused on collecting usable aerial imagery while identifying and responding to challenges such as software issues, environmental hazards, and unexpected interactions on site. Through these missions, we gained hands-on experience with Skydios autonomous mapping capabilities and reinforced the importance of preparation, adaptability, and proper communication during UAS operations.

Red Rectangle (Lawnmower Grid) Mission

Table 24

Table 24

Altitude

200 feet

Overlap

80% frontal and side

Lens Angle

90° (Nadir)

Flight time

approximately 20 minutes.

There are no rows in this table

⁠

Blue Rectangle-Size Area (Cross Hatch + Perimeter)

Table 25

Table 25

Altitude

60–80 feet ( for clearance of obstacles)

Overlap

80%

Lens Angle

75° bevel

There are no rows in this table

⁠

Extra Notes

Table 26

Table 26

Location

Weather

Hazards

Notes

Student garden

Sunny

Trees, gardeners

One pedestrian interaction.

There are no rows in this table

⁠

DSM Red boundary

⁠

Figure 6 DSM red boundary

Table 27

Table 27

Cell Size

X= 2.782cm Y= 2.782cm

Pixel Depth

32 bit

DSM Value

205.072-237.771

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Review for DSM

The DSM effectively shows height differences between structures, vegetation, and open ground. Taller objects such as buildings stand out clearly against flatter terrain. Some elevated noise appears along tree lines, which is expected due to irregular vegetation and shadows impacting point matching.

Shaded DSM red boundary

⁠

Figure 7 Shaded DSM of purdue student garden

Review for shaded DSM of Red boundary

The DSM clearly captures surface features such as greenhouses, buildings, and taller vegetation. Rectangular patterns from the garden plots and structures are easy to identify, which indicates good image overlap and flight consistency. Slight noise appears near the outer edges and along taller objects, likely caused by shadows and angle changes during image capture.

Table 28

Table 28

Cell Size

X= 2.782cm Y= 2.782cm

Pixel Depth

32 bit

DSM Value

205.072-237.771

Shaded DSM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DSM blue

⁠

Figure 8 shaded dsm blue

Table 29

Table 29

Cell Size

X= 2.782cm Y= 2.782cm

Pixel Depth

32 bit

DSM Value

205.072-237.771

Shaded DSM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DTM blue boundary Review

This visualization helps emphasize elevation patterns while still showing underlying imagery. Flight lines and image centers appear evenly spaced, showing good mission planning. Any slight inconsistencies in height values occur mostly near the mission boundary, again pointing to reduced overlap at the edges.

DTM blue boundary

⁠

Figure 9DTM blue boundary

Table 30

Table 30

Cell Size

X= 1.391cm Y=1.391cm

Pixel Depth

32 bit

DTM Value

205.134– 209.143

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

DTM review

This DTM looks good overall and clearly represents the underlying terrain without buildings or vegetation height. Elevation changes are smooth, especially across open grassy areas, which shows that the terrain filtering worked correctly. Some edge areas appear slightly smeared or stretched, which is likely due to less overlap near the boundary of the flight area. These minor distortions relate to the fact that data density drops off at the edges of the mission.

Shaded DTM blue boundary

⁠

Figure 10DTM Blue boundary

Table 31

Table 31

Cell Size

X= 1.391cm Y=1.391cm

Pixel Depth

32 bit

DTM Value

205.134– 209.143

Shaded DTM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DTM blue boundary review

This DTM does a good job removing above ground features and showing elevation trends across the site. The center of the map looks very clean, while the perimeter shows small elevation inconsistencies. These issues likely come from fewer image tie points near the edges and slight elevation smoothing during processing.

Orthomosaic blue

⁠

Figure 11Orthomosaic blue boundary

Table 32

Table 32

Cell Size

X= 2.782 Y = 2.782cm

Pixel Depth

32 bit

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Orthomosaic Map review

This final orthomosaic clearly combines both boundary areas and shows consistent alignment between datasets. The POIs are sharp and well-defined, especially in the central overlap region. Minor distortion near the outer boundary is visible but does not significantly impact usability. These small issues align with field conditions such as changing lighting and the limits of a single-battery mission.

Ortho mosaic RED boundary

⁠

Figure 12orthomosaic red boundary

Table 33

Table 33

Cell Size

X= 2.782 Y = 2.782cm

Pixel Depth

32 bit

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Orthomosaic Review Red boundary

The orhomosaic is visually strong, with good color balance and sharp detail in the center of the map. Garden beds, roads, and building are clearly visible. The insets highlight areas of interest well, although minor stretching and blur are noticeable near the boundary edges. This relates to the flight path turning points where overlap can briefly decrease.

Data Set 3 M300 Sensor Purdue Turf farm

This lab focused on planning and completing a thermal mapping mission using the DJI Matrice 300 equipped with the H20T sensor. The objective was to collect high-overlap thermal and RGB imagery while also deploying a GNSS static base station to support accurate georeferencing. Due to crowd activity from a nearby pesticide presentation, the team relocated to a safer setup area, but weather conditions remained ideal and allowed the mission to be completed successfully. The flight highlighted the need for careful mission planning, teamwork, and flexibility in the field, and the resulting dataset will be used in future labs for thermal analysis, ground control accuracy, and mapping workflow evaluation.

DSM

⁠

Figure 13 DSM of m300

Table 34

Table 34

Cell Size

X= 5.839cm Y= 5.839cm

Pixel Depth

32 bit

DSM Value

175.118-197.406

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

DSM Review m300

The DSM highlights elevated features such as vehicles and structures clearly, showing strong height differentiation across the scene. The clustered high-elevation values near the parking area match real-world features observed during data collection. Some exaggerated elevation spikes appear near the right side of the map, which may be caused by thermal contrast differences affecting point matching. This reflects the sensor limitations more than poor mission planning.

Shaded DSM

⁠

Figure 14Shaded DSM m300

Table 35

Table 35

Cell Size

X= 5.839cm Y= 5.839cm

Pixel Depth

32 bit

DSM Value

175.118-197.406

Shaded DSM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DSM Review m300

The shaded DSM clearly shows surface features such as vehicles, buildings, and vegetation as elevated objects above the ground. The hillshade effects enhances texture and makes man made objects stand out well, especially along the eastern side where parked vehicles are visible. A consistent grid pattern from the flight lines indicates good overlap and stable flight paths. Some roughness and distortion appear near the edges of the dataset, which is typical and likely caused by reduced overlap at the boundary of the mission area. These edge artifacts relate to normal data collection limitations rather than major flight errors

DTM

⁠

Figure 15 DTM M300

Table 36

Table 36

Cell Size

X= 2.919cm Y=2.919cm

Pixel Depth

32 bit

DTM Value

176.482-183.077

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

DTM Review m300

This DTM emphasizes overall elevation trends across the turf farm without surface clutter. The interior of the dataset appears uniform and reliable, which shows strong alignment and processing. Warping and exaggerated elevation values near the boundaries indicate areas where fewer images contributed to the model. These artifacts likely relate to the limited usable texture in thermal imagery rather than flight execution issues.

Shaded DTM

⁠

Figure 16Shaded DTM M300

Table 37

Table 37

Cell Size

X= 2.919cm Y=2.919cm

Pixel Depth

32 bit

DTM Value

176.482-183.077

Shaded DTM Value

0-255

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Shaded DTM Review m300

The shaded DTM does a good job removing surface objects and representing the bare-earth terrain. Elevation changes appear smoother and more gradual compared to the DSM, which indicates successful filtering of buildings and vehicles. Slight elevation noise is noticeable near the outer edges of the map, suggesting less reliable data where image coverage thins. Overall, the consistent terrain surface suggests the flight altitude and overlap were sufficient for terrain modeling despite using a thermal sensor.

Orthomosaic

⁠

Figure 17Orthomosaic M300

Table 38

Table 38

Cell Size

X= 2.782 Y = 2.782cm

Pixel Depth

32 bit

Coordinate Projection System

WGS 1984 UTM Zone 16N Transverse Mercator EGM96 Height (current Z)

There are no rows in this table

⁠

Orthomosaic Review M300

The thermal Ortho mosaic successfully captures temperature differences across the site, with vehicles and hard surfaces appearing warmer than surround turf. Patterns such as tire tracks, parked vehicles, and surface texture are clearly visible, showing good image alignment. Some areas appear blurred or less sharp, especially In uniform grass regions, which is expected with thermal sensors due to lower spatial resolution. These issues are consistent with sensor characteristics rather than overlap or flight path problems.

Conclusion

Overall, AT 309 helped me better understand how UAS data collection and GIS processing work together to create useful mapping products. Over the summer, I learned how raw imagery from different aircraft and sensors can be turned into orthomosaics, DSMs, DTMs, and shaded elevation models using software like Drone2Map and ArcGIS Pro. Working with multiple platforms, including the DJI Mavic 2, Skydio 2+, and DJI Matrice 300, showed me how different flight patterns, altitudes, and sensors affect data quality and final outputs.

One thing I appreciated most was seeing how much flight planning and field decisions matter. Issues like weather, obstacles, crowd activity, and overlap settings directly impacted the quality of the data, and having to adapt in the field made the labs feel realistic. I also gained a better understanding of coordinate systems, projections, and why consistency is important when comparing datasets. Processing multiple datasets with similar workflows made it easier to recognize patterns in data quality and understand why some areas had errors or lower accuracy.

Overall, this class helped connect flying drones with real-world GIS applications. AT 309 gave me hands-on experience that improved both my technical skills and my confidence working with spatial data, and it helped me see how UAS mapping can be used for research, planning, and analysis beyond the classroom.

Want to print your doc?
This is not the way.

Try clicking the ··· in the right corner or using a keyboard shortcut (

CtrlP

) instead.