Skip to content
Week 6 Thermal Mapping Mission
1. Introduction
This lab focused on creating and executing a thermal mapping mission using a DJI Marvic 300 (M300) equipped with the Zenmuse H20T sensor. The H20T collects imagery in both RGB and thermal infrared (IR), enabling multi-modal documentation of surface features and temperature patterns. In lab, we created the mission on the DJI controller using the DJI Pilot app, with additional discussion of how a mission could be designed in Google Earth and imported to the controller.
The two primary objectives of this exercise were to discover, identify, and implement the creation of a mapping mission using the M300 with a thermal sensor and understand the fundamentals of ground control by setting up a GNSS static base for later use. Because thermal imagery generally has lower effective bit depth and can be more sensitive to noise and subtle radiometric variation, a higher overlap was required compared to standard RGB mapping. Therefore, the mission was planned with 90% frontal and 90% lateral overlap at 60 meters AGL with a 90° nadir camera orientation.
2. Study Areas
Two sub-areas within the Purdue Turf Farm complex were used for this lab: the Student Garden area and the Turf Parking area. Both locations provide accessible, open environments with a mix of turf, paved surfaces, and nearby structures that support evaluation of thermal mapping workflows and the effects of high overlap on dataset quality.
The Student Garden area included maintained plots, adjacent walkways, and localized surface material changes that are useful for observing thermal contrast between vegetation, soil, and hardscape. The Turf Parking area included larger paved zones and surrounding turf, offering a complementary setting to observe thermal differences across broad uniform surfaces and edge transitions.
Figure 1. M300 staged on a designated landing pad at Purdue Turf Farm prior to mission execution.
3. Mission Planning
3.1 Mission Creation on DJI Pilot 2
We created a parallel ‘lawn mower’ grid mission in the lab using the DJI Pilot app on the controller. The grid parameters were set to 60 m AGL with a 90% frontal and 90% lateral overlap to meet thermal mapping requirements. A 90° nadir orientation was selected to maintain consistent geometry for photogrammetric and thermal mosaic processing. The mission polygons were reviewed to ensure they covered only the assigned blue area for each location.
Where relevant, the team reviewed how a comparable polygon and grid could be drafted in Google Earth and imported into the controller to streamline repeatable mission planning across sites.
Figure 2. DJI Pilot 2 in-flight interface showing mission execution with thermal/RGB view during a grid route.
3.2 Pre-Flight Weather Review
A pre-flight weather review was conducted for the KLAF area. The METAR indicated light winds and clear conditions around the time of operations (e.g., variable winds near 6 knots, 10 statute miles visibility, clear skies). These benign conditions supported stable grid tracking and reduced the risk of motion blur. Such stability is particularly valuable for high-overlap thermal missions that depend on consistent line spacing and uniform sampling.
Figure 3. AWC weather snapshot for KLAF used for pre-flight go/no-go and documentation of conditions.
4. Gathering the Data
In the field, the flight crew confirmed roles (Pilot in Command and Visual Observer) and verified that the mission settings matched the thermal requirements. We established a clear takeoff and landing zone and maintained situational awareness of people, vehicles, and equipment in the vicinity of the operating area. Throughout both missions, we monitored the aircraft’s adherence to the grid route and verified that the H20T sensor was capturing RGB and thermal imagery as expected.
Figure 4. M300 in flight during thermal grid data collection over the Purdue Turf Farm area.
4.1 Site Conditions
Both the Student Garden and Turf Parking areas were easy to access with sufficient open space for safe launch and recovery. Pedestrian activity was light and manageable, with the majority of movement limited to occasional passersby near pathways and parking edges. The terrain was mostly flat, and the open layout reduced the likelihood of unexpected line-of-sight obstructions during the automated grid flights.
Because thermal data quality can be influenced by surface heating and diurnal timing, we paid attention to the general thermal state of turf, pavement, and garden plots at the time of capture. The mid-day conditions provided sufficient thermal contrast for qualitative interpretation while maintaining consistent lighting for the paired RGB dataset.
4.2 Potential Hazards and Mitigation
Potential hazards across the two sites included: (1) intermittent vehicle movement in and near the Turf Parking area, (2) localized pedestrian presence near the Student Garden and adjacent walkways, (3) trees, light poles, and minor structures at the perimeter that could affect obstacle margins, (4) higher mission workload due to the 90%/90% overlap requirement, and (5) possible GNSS multipath near built features.
Mitigation strategies included establishing a marked safety buffer around the takeoff/landing zone; maintaining continuous two-person communication; pausing or re-positioning operations if vehicles or pedestrians approached the immediate area; verifying altitude margins and obstacle avoidance settings; and conducting deliberate pre-mission checks of overlap and sensor mode. For GNSS activities, the base station was placed in an open-sky location to minimize multipath and ensure stable static observations.
5. GNSS Base Station and Static Data Collection
In addition to the aerial missions, we set up a GNSS base station and collected static data to support later use as a reference base. The setup process included selecting an open-sky location, leveling and securing the tripod or mounting platform, recording antenna height measurements, and initiating a static observation session of sufficient duration for post-processing. This activity reinforced the role of ground control and reference infrastructure in improving positional accuracy for mapping workflows and future labs.
6. Data Review, Metrics, and Storage
Following each mission, imagery was reviewed to confirm that files were recorded correctly and that the grid coverage appeared complete. The high-overlap configuration produced large, well-sampled image sets appropriate for thermal mosaicking. The datasets were saved to dedicated storage and shared with each flight crew member as required for future lab activities.
6.1 Student Garden Mission Metrics
Table 6
Metric
Recorded Value
Study Area
Purdue Turf Farm – Student Garden
Time to drive to the field and set up
30 minutes
Time to fly the mission
30 minutes
Number of images captured
662 images
Average size of each image
≈ 1.3 MB per image
Total dataset size
0.9 GB
There are no rows in this table
The high image count aligns with the 90%/90% overlap requirement, which is intended to compensate for the relative radiometric limitations of thermal imagery and improve the consistency of the final thermal mosaic.
6.2 Turf Parking Mission Metrics
Table 7
Metric
Recorded Value
Study Area
Purdue Turf Farm – Turf Parking
Time to drive to the field and set up
30 minutes
Time to fly the mission
30 minutes
Number of images captured
796 images
Average size of each image
≈ 1.4 MB per image
Total dataset size
1.1 GB
There are no rows in this table
7. Data Collection Outcomes / Expected Deliverables
Across the two sites, the M300 + H20T missions produced paired RGB and thermal datasets suitable for creating RGB mosaucs and thermal mosaics. The Student Garden dataset is expected to support interpretation of thermal variation across vegetation, soil, and small-scale surface features, while the Turf Parking dataset provides a complementary view of thermal behavior over large paved surfaces and adjacent turf. The GNSS static base data collected during this lab will be used in subsequent workflows to improve georeferencing accuracy and reinforce fundamentals of ground control.
Figure 5. Example H20T thermal frame showing temperature contrast across built and vegetated features in the Turf Farm area.
Figure 6. Additional thermal frame captured during grid operations to support later mosaic generation and analysis.
8. DJI Pilot Workflow Notes and Observations
DJI Pilot provided an integrated workflow for defining the mission polygon, assigning a parallel grid pattern, configuring altitude and overlap, and verifying sensor mode prior to launch. The interface supported real-time monitoring of route execution and confirmation that the H20T thermal feed and RGB capture were active.
A key operational consideration for thermal mapping is that the 90% overlap requirement increases route length and image volume, which can affect battery planning and mission pacing. Structured pre-flight checks and clear Pilot–Observer coordination help ensure that overlap settings, altitude, and sensor mode are not overlooked.
9. Conclusion
This lab demonstrated the planning and execution of thermal mapping missions using the DJI M300 equipped with the H20T sensor across two Purdue Turf Farm sub-areas: Student Garden and Turf Parking. Missions were flown using a parallel grid pattern at 60 m AGL and configured with 90% frontal and 90% lateral overlap to address thermal imagery characteristics and to support robust mosaic generation.
The Student Garden mission produced 662 images (0.9 GB) and the Turf Parking mission produced 796 images (1.1 GB), each with a 30-minute drive/setup period and a 30-minute flight duration. In parallel, the GNSS static base station setup reinforced ground control fundamentals and will provide a reference dataset for accuracy improvements in subsequent processing. Overall, the combined aerial and GNSS activities aligned well with the stated lab objectives and provided a strong foundation for future deliverables.
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.