Skip to content
Final Data Product Assignment
Introduction
Over the course of this semester, my team and I developed a strong foundation in unmanned aerial systems mapping by conducting a series of field missions and transforming collected data into clear, accurate geospatial products. Using multiple aircraft platforms Skydio 2+, DJI Mavic 2 Pro, and DJI Matrice 300 we captured imagery and flight metadata across a range of operational scenarios. Most missions were completed at the Purdue University Student Garden and Turffarm. These sites provided realistic environments for practicing mission planning, safe field operations, and repeatable data collection.
Throughout the labs, we flew mapping patterns such as 3D scans, perimeter scans, and crosshatch scans, allowing us to compare how flight geometry, overlap strategies, and mission design influence model quality. The datasets produced from these missions included essential information such as GPS coordinate references, mission time and date, camera settings, and flight logs, which together support reliable mapping, 3D reconstruction, and spatial analysis.
To convert raw imagery into actionable outputs, I used Drone2Map for photogrammetric processing and ArcGIS Pro (with support from tools like ArcGIS Earth) to create professional cartographic products. This final report represents the culmination of that workflow: organizing and processing mission data, validating results, and presenting readable and accurate 2D and 3D map products. By working across three distinct drone platforms, each with different strengths in autonomy, imaging performance, and positional accuracy, I gained practical insight into how aircraft and sensor capabilities directly affect mission outcomes. Overall, this project demonstrates my ability to move from field collection to final map production using industry-relevant methods and GIS tools.
Processing
For all three datasets, I used the same processing workflow. First, I organized each mapping mission into separate, clearly labeled folders. This kept the datasets clean and easy to manage. Next, I imported the imagery into Drone2Map. I reviewed the basic project settings before processing. I then selected the standard 2D outputs: Orthomosaic, DSM, and DTM. After confirming the output selections, I started processing and monitoring the steps. Drone2Map completed image alignment, dense matching, and final product generation. When processing finished, Drone2Map produced the orthomosaic and elevation models in a 2D map view.
After Drone2Map processing, I exported the final outputs and brought them into ArcGIS Pro. In ArcGIS Pro, I created five maps for each dataset. These were a DSM map, a DTM map, shaded DSM, shaded DTM, and an orthomosaic map. The DSM shows the full surface, including buildings and vegetation. The DTM shows the bare-earth surface by removing above-ground features. I used hillshade to create shaded DSM and shaded DTM. This improved readability and made elevation changes easier to interpret. The orthomosaic provided a geometrically corrected image for accurate visual assessment and measurements.
For each mission, Drone2Map also produced processing details. These included cell size, pixel depth, elevation ranges, and coordinating projection information. Keeping the workflow consistent allowed me to compare results between different flight patterns and conditions. Most settings stayed the same across missions. Overall, this workflow showed that good data organization and consistent processing produce reliable GIS-ready products.
DJI Mavic 2 Pro
• Altitude: 60–80 ft
• Camera: Oblique (~75°)
• Crosshatch and perimeter enabled
• During flight, GPS signal was lost, requiring manual intervention (Take Control). After relaunch, the scan continued, reaching 268 of 375 photos captured at 80 ft. Altitude: 200 ft
• Camera: Nadir (90°)
• Overlap: 80% side / 80% front
• Result: 67 images, 9 min flight, no cross-hatch or perimeter enabled.
LAANC approval confirmation via SMS
Detailed weather: RH 47%, no clouds, Kp=1.
METAR summary: VFR, winds 150°@5kt, 21.7°C.
Figure 1: DSM Red Box
Table 8
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
1.65cm
Cell Size Y
1.65cm
Pixel Depth
32 bit
DSM Value
202.344 - 246.642
There are no rows in this table
The changes in surface height are easy to see, especially around buildings and vegetation.
Some areas look patchy or speckled, which suggests the elevation values are less accurate there.
This likely happened because the image overlap was lower and the flight did not include cross hatch paths,
Overall, the DSM is usable, but it looks rougher and less detailed than a full grid mission DSM.
Figure 2: Shaded DSM Red Box
Table 9
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
1.65cm
Cell Size Y
1.65cm
Pixel Depth
32 bit
DSM Value
202.344 -246.642
Shaded DSM Value
0-255
There are no rows in this table
The DSM shows surface features clearly, like buildings, tree lines, and changes in vegetation height.
The flight lines and image centers look consistent, which suggests good coverage and overlap.
Higher elevation values along the west side are likely from trees and uneven vegetation.
Near the boundary edges, the DSM looks a bit noisy and rough. This is probably because overlap is weaker at the edges, which is common in boundary-only flights.
Figure 3: DTM Red Box
Table 10
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
8.2337 cm
Cell Size Y
8.2337 cm
Pixel Depth
32 bit
DSM Value
202.419 – 222.546
There are no rows in this table
The terrain surface looks smoother than the DSM, so it shows the overall ground shape more clearly.
Changes in elevation look more even across the area.
Some edge areas look distorted, likely because there were not enough good ground points to classify correctly.
This shows the limits of flying only the boundary instead of covering the whole site.
Figure 4: Shaded DTM Red Box
Table 11
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
8.2337 cm
Cell Size Y
8.2337 cm
Pixel Depth
32 bit
DSM Value
202.419 – 222.546
Shaded DSM Value
0-255
There are no rows in this table
DTM removes most objects above the ground, so the terrain looks smoother.
You can see the ground slowly getting higher or lower from west to east, which shows the natural slope.
A few rough spots near the west edge are still there. This is likely because thick vegetation made the ground harder to see in the images.
This kind of issue is common when you fly only the boundary instead of flying a full grid inside the area.
Figure 5: True Ortho Red Box
Table 12
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
1.65cm
Cell Size Y
1.65cm
Pixel Depth
32 bit
There are no rows in this table
The orthomosaic looks clear, with consistent color and sharp detail.
Buildings, roads, and field patterns match well with real-world locations.
There is a small amount of stretching or distortion near the outer edges.
These edge problems are common when overlapping is limited, and the flight only covers the boundary.
In this lab, our team flew two mapping missions with a Skydio at the Purdue Student Garden. The goal was to practice planning, flying, and managing autonomous mapping flights.
This helped us simulate real mapping work. We focused on collecting good aerial images and dealing with common problems like software errors, environmental hazards, and unexpected situations on site. Overall, this lab gave us real hands-on experience with Skydio’s autonomous mapping and showed how important preparation, flexibility, and clear communication are during UAS operations.
Sky Dio 2 +
Study Area: Purdue Student Garden
Platform/UAS: Sky Dio 2 +
Flight App: DroneDeploy Flight App
Mission Type: Parallel ‘lawn mower’ grid mapping
Planned Altitude: 60-80 fts/ 200fts
Camera Orientation: 90° nadir (straight down)
Overlap Target: 80% frontal / 80% lateral
DroneDeploy mission planning view showing the mapping polygon and parallel grid layout over Purdue Student Garden.
LAANC Authorization Request pre-check showing auto-approval eligibility under LAF in Class D airspace.
Pre-flight weather summary for the KLAF area indicating VFR conditions and light winds.
A pre-flight weather review was conducted using an aviation weather summary tool for the KLAF area. Conditions were reported as VFR with no warnings. Visibility was 10 miles with few clouds, and surface winds were light at approximately 4 knots from roughly 160°. The temperature was approximately 22.2°C. These conditions were considered favorable for mapping due to reduced turbulence and a lower risk of motion blur, supporting consistent image capture for high-overlap photogrammetry.
Figure 6: DSM Red Box
Table 13
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
DSM Value
205.072-237.771
There are no rows in this table
The DSM clearly shows height differences between buildings, vegetation, and open ground. Taller features like buildings stand out from the flatter areas. There is some extra “noise” along the tree lines, which is normal because trees and shadows can make matching points harder.
Figure 7: Student Garden Shaded DSM
Table 14
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
DSM Value
205.072-237.771
Shaded DSM Value
0-255
There are no rows in this table
The DSM clearly shows features like greenhouses, buildings, and taller plants. The garden plots and structures are easy to see, which suggests the flight had good overlap and consistent coverage. There is a little noise near the edges and around tall objects, likely because shadows and camera angle changes can affect the images.
Figure 8: Student Garden Shaded DSM Blue Box
Table 15
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
DSM Value
205.072-237.771
Shaded DSM Value
0-255
There are no rows in this table
This view makes it easier to see elevation patterns while still showing the image underneath. The flight lines and image centers are evenly spaced, which suggests the mission was planned well. Small height errors mostly show up near the edges of the mission, likely because overlap is lower at the boundary.
Figure 9: Student Garden Shaded DTM Blue Box
Table 16
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
1.49cm
Cell Size Y
1.49cm
Pixel Depth
32 bit
DSM Value
205.134– 209.143
There are no rows in this table
This DTM looks good overall and shows the ground shape clearly, without building or vegetation height. Elevation changes are smooth, especially in open grassy areas, which suggests the ground filtering worked well. Some areas near the edges look slightly smeared or stretched, likely because there was less overlap at the boundary. This happens because there are fewer data points near the edges of the mission.
Figure 10: Student Garden Shaded DTM Blue Box
Table 17
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
1.49cm
Cell Size Y
1.49cm
Pixel Depth
32 bit
DSM Value
205.134– 209.143
Shaded DSM Value
0-255
There are no rows in this table
This DTM removes most above-ground objects well and shows the overall elevation changes across the area. The center looks very clean, but the edges have small height errors. This is likely because there are fewer matching points near the boundary and the software smooths the surface a bit during processing.
Figure 11: Student Garden Orthomosaic Blue Box
Table 18
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
There are no rows in this table
Orthomosaic combines both boundary areas well, and the two datasets line up consistently. Key features (POIs) look sharp, especially in the center where the images overlap the most. There is a little distortion near the outer edge, but it does not affect the map much. These small issues are normal and may be due to changing light and the limits of flying the mission on one battery.
Figure 12: Student Garden Orthomosaic Red Box
Table 19
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
There are no rows in this table
The orthomosaic looks clear overall, with good color and sharp detail in the middle of the map. Garden beds, roads, and buildings are easy to see. The inset maps show the key areas well, but there is a little stretching and blur near the outer edges. This is likely because overlap can drop slightly when the drone turns at the end of the flight lines.
M300
Study Area: Purdue Turffarm
Platform/UAS: M300
Mission Type: Parallel ‘lawn mower’ grid mapping
Planned Altitude: 60 meter
Camera Orientation: 90° nadir (straight down)
Overlap Target: 90% frontal / 90% lateral
n this lab, we planned and flew a thermal mapping mission using a DJI Matrice 300 with the H20T sensor. Our goal was to collect thermal and RGB images with high overlap. We also set up a GNSS static base station to improve georeferencing accuracy.
Because there was a crowd nearby for a pesticide presentation, we moved to a safer setup location. The weather was good, so we were still able to complete the mission successfully.
This flight showed the importance of good planning, teamwork, and being flexible in the field. We will use this dataset in future labs for thermal analysis, checking ground control accuracy, and evaluating the mapping workflow.
Figure 13: M300 DSM
Table 20
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
5.84cm
Cell Size Y
5.84cm
Pixel Depth
32 bit
DSM Value
175.118-197.406
There are no rows in this table
The DSM clearly shows taller features like vehicles and buildings, with clear height differences across the area. The higher elevation values near the parking lot match what we saw in the field. A few unusually high spikes show up on the right side of the map, likely because thermal contrast can make point matching harder. This seems more related to sensor limits than to the flight plan.
Figure 14: M300 Shaded DSM
Table 21
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
5.84cm
Cell Size Y
5.84cm
Pixel Depth
32 bit
DSM Value
175.118-197.406
Shaded DSM Value
0-255
There are no rows in this table
The shaded DSM clearly shows raised features like vehicles, buildings, and vegetation above the ground. The hillshade adds texture and helps man-made objects stand out, especially on the east side where parked cars are easy to see. The flight lines form a consistent grid, which suggests good overlap and a stable flight. Some rough or distorted areas appear near the edges, which is normal and likely caused by lower overlaps at the boundary. These edge effects are common and are due to normal data limits, not major flight errors.
Figure 15: M300 DTM
Table 22
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.92cm
Cell Size Y
2.92cm
Pixel Depth
32 bit
DSM Value
176.482-183.077
There are no rows in this table
This DTM shows the overall ground elevation across the turf farm without extra surface detail. The middle of the map looks smooth and reliable, which suggests the data aligned and processed well. Near the edges, some areas look warped and the elevation values seem too high or too low, likely because fewer images were used there. This is probably due to thermal images having less clear texture to match, not because of problems with the flight.
Figure 16: M300 Shaded DTM
Table 23
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2.92cm
Cell Size Y
2.92cm
Pixel Depth
32 bit
DSM Value
176.482-183.077
Shaded DSM Value
0-255
There are no rows in this table
The shaded DTM removes most surface objects and shows the bare ground well. Elevation changes look smoother than in the DSM, which means buildings and vehicles were filtered out successfully. A little elevation noise appears near the outer edges, where image coverage is weaker. Overall, the terrain looks consistent, which suggests the flight altitude and overlap were good enough for terrain mapping even with a thermal sensor.
Figure 17: M300 Ortho Mosaic
Table 24
Projected Coordinate System
WGS 1984 UTM Zone 16N
Projection
Transverse Mercator
Cell Size X
2,78cm
Cell Size Y
2.78cm
Pixel Depth
32 bit
There are no rows in this table
The thermal orthomosaic shows temperature differences clearly. Vehicles and hard surfaces look warmer than the surrounding turf. Features like tire tracks, parked cars, and surface texture are easy to see, which suggests the images lined up well. Some areas look a bit blurry, especially over uniform grass, which is normal for thermal sensors because they have lower detail (resolution). These issues are mainly due to the sensor, not because of poor overlap or flight path planning
Conclusion
Overall, AT 309 helped me understand how UAS data collection and GIS processing work together to create useful maps. I learned how raw images from different drones and sensors can be processed into products like orthomosaics, DSMs, DTMs, and shaded elevation models using Drone2Map and ArcGIS Pro. Using different platforms—DJI Mavic 2, Skydio 2+, and DJI Matrice 300—also showed me how flight patterns, altitude, and sensors can change data quality and results.
The most valuable parts of the class was learning how important flight planning and field decisions are. Things like weather, obstacles, people nearby, and overlap settings affected the data quality a lot. Having to adjust in the field made the labs feel more like real work. I also learned more about coordinate systems and map projections, and why it matters to keep them consistent when comparing datasets. Processing multiple datasets with similar steps helped me notice common error patterns and understand why some areas had lower accuracy.
This class connected drone flying with real GIS applications. AT 309 gave me practical experience, improved my technical skills, and made me more confident working with spatial data. It also helped me see how UAS mapping can be used for research, planning, and analysis outside of class.
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ··· in the right corner or using a keyboard shortcut (
CtrlP
) instead.