5 Settings to Improve the Accuracy of Drone Surveying

Table of Contents

• How much can the accuracy of drone surveying change with settings?

• Setting 1 Adjust the shooting altitude to suit the purpose

• Setting 2: Ensure sufficient overlap

• Setting 3 Fix the shutter speed and exposure conditions

• Setting 4: Adjust flight speed and shooting interval

• Setting 5: Review reference points and position correction settings

• Field factors other than settings that commonly cause accuracy differences

• Summary

How much can drone survey accuracy be changed by settings?

When aiming to improve the accuracy of drone surveying, many field personnel initially focus on the aircraft's performance and differences in processing software. Of course, those factors also affect the deliverables. However, in practice, even when using the same aircraft and the same processing environment, the finished results can often differ depending simply on how settings are fine-tuned. In particular, on land development sites with undulating terrain, construction sites with complex structures, and civil engineering sites where slopes and temporary structures coexist, insufficient initial settings tend to lead directly to re-surveys and rework.

Accuracy in drone surveying can be viewed from several perspectives. These include how well planimetric positions align, how consistent elevations are, how well point clouds and orthophotos stitch together, the reproducibility required for as-built verification, and the ease of comparing with past data. On site, simply 'looking good' is not enough; the data must be usable for drafting, earthwork quantity verification, progress comparison, construction records, and explanations to stakeholders. Therefore, it is important to produce data that is not only visually pleasing but also numerically stable and resistant to fluctuation.

What you should note here is that accuracy is not determined by a single setting. Lowering the altitude will capture finer detail, but it increases the number of flights and photos taken, raising the processing load. Increasing the overlap will stabilize image alignment, but depending on site conditions it can needlessly increase data volume. Even if you increase shutter speed to reduce blur, shooting in low light or under overcast skies can cause sensitivity settings to fluctuate and introduce other sources of error. In other words, settings do not act alone; they interact and collectively influence overall accuracy.

What operational personnel need to grasp is the perspective of identifying which settings in their current procedures to revise to stabilize results before adding costly systems. It is not uncommon for settings decided in the few minutes before a flight to greatly influence subsequent analysis results. Conversely, if you can organize your approach to settings to suit field conditions, there is ample room to improve accuracy even with the same equipment.

This article focuses on five settings that typically produce the greatest differences in drone surveying accuracy. Rather than merely explaining terms, it organizes, along the practical workflow, why each setting affects accuracy, in what field conditions failures are likely to occur, and how to decide which option to choose. It will be useful not only for personnel who want to review their operations but also for those who need to assess the quality of outsourced results.

Setting 1: Adjust the shooting altitude to match the purpose

One of the fundamental settings that affects the accuracy of drone surveying is the flight altitude. Flight altitude directly determines how finely the ground is captured. If the altitude is too high, you can photograph a wide area in a short time, but the ground representation per image becomes coarse and the reproducibility of small steps and boundaries tends to decline. Conversely, lowering the altitude makes it easier to capture details, but increases the number of flights and images, raising processing time and the burden of data management. Therefore, it is not simply a matter of lower being better; it needs to be optimized according to the objective.

For example, if the goal is to obtain an overall understanding of a large development site, an approach that prioritizes the continuity and efficiency of the entire surface is effective. On the other hand, if you need soil volume calculations, verification of as-built conditions, locating the slope shoulder and slope toe, or checking the geometry around structures, coarse capture settings tend to reveal shortcomings afterward. A common mistake when trying to improve accuracy is raising the flight altitude too much because of an excessive desire to shorten flight time. On site, the relief of having completed the capture comes first, but if processing later shows poorly defined lines, collapsed edges, or missing fine undulations, re-surveying will ultimately be the greater loss.

When deciding imaging altitude, it's important to first clarify what you want as the deliverable. Whether the focus is on orthophotos, whether you'll also use point clouds and 3D models, or whether you'll carry out comparisons with drawings after completion will change the required level of detail. Identifying in advance which features you'll want to check later—not just the ground surface but also embankment shoulders, excavation edges, edges of side ditches or gutters, pavement transitions, and areas around temporary enclosures—will help keep altitude-setting decisions consistent.

Also, if you ignore elevation differences and fly at a uniform altitude, the effective resolution will vary by location. On sites with significant terrain variation, lower areas may be captured in finer detail while higher areas become relatively coarser. In such sites, planning flights with the terrain in mind produces more consistent results than applying a single uniform setting across the whole area. At sites that include slopes or earthworks with large steps, taking care not only with simple planar imaging but also to ensure that the distance to the target surface does not vary greatly directly leads to improved accuracy.

Furthermore, flight altitude must be balanced with safety. At sites where there are power lines, trees, temporary installations, cranes, or existing structures nearby, low-altitude flight is not necessarily advantageous. Rather than lowering altitude at the expense of safety margins, it is practical to seek a height that avoids hazards while ensuring the required accuracy. Pursuing accuracy alone at the cost of operational safety will not result in reproducible operations.

An altitude setting for improving accuracy is not simply flying lower; it means matching the height so the subject can be captured stably at the required level of detail. Since each site differs in size, terrain undulation, obstacles, and desired outcomes, it is important not to simply reuse the settings that worked last time, but to reassess each time "how much detail is needed for this site." Simply adopting this mindset can significantly reduce variability in the final results.

Setting 2: Ensure sufficient overlap

In stabilizing the accuracy of drone surveys, a frequently overlooked but critically important factor is overlap — the amount of overlap between photos. If overlap in the along-track or cross-track directions is insufficient, the connections between images weaken and alignment becomes unstable. As a result, issues such as distortion of the entire model, collapse at the edges, local height shifts, and disturbances around structures are likely to occur. Even if it appears that many photos were captured on site, inadequate overlap will not lead to highly accurate results.

What's particularly common in practice is using settings that were fine for flat ground exactly as they are at a different site. Even ground surfaces that look monotonous, such as development sites, can make it difficult for the image-processing side to pick up feature points when the soil appearance is uniform and there are few landmarks. Furthermore, when slopes, fill edges, puddles, material yards, temporary roads, and the like coexist, the ease of connection varies by location. If there isn't enough margin in the overlap rate, local errors can more easily affect the entire result.

The purpose of increasing overlap is not simply to increase the number of photos. It is to improve the stability of three-dimensional reconstruction by capturing the same ground surface or object multiple times from slightly different positions. If overlap is sufficient, even if some photos have blur or variations in brightness, the dataset as a whole is more likely to compensate. Conversely, when overlap is low, the processing results become highly dependent on the quality of each individual photo, and small errors can lead to large shape distortions.

However, even in this case, simply raising it to the maximum is not the right approach. The more you increase overlap, the more photos are taken, which increases the load on flight time, battery consumption, processing time, and storage capacity. In practice, it is important to secure sufficient overlap that effectively improves accuracy while keeping it within operationally feasible limits. You need to decide how much margin to allow based on the size and terrain of the site, the complexity of the subject, the strength of the wind, and how pronounced the surface features are.

At sites with many structures or large elevation differences, shooting only from directly overhead often fails to capture enough information, and even when overlap rates are met, reconstruction of the sides can be weak. In such cases, rather than looking only at numerical overlap, it is important to plan flights by considering which faces are being captured and where blind spots will occur. Improving accuracy depends not only on how much the photos overlap each other but also on how many directions the subject is captured from.

Also, the importance of overlap grows on windy days. Even a slight drift of the aircraft can pull it off the planned path, causing the actual overlap to be less than expected. Even if your settings look safe on paper, anticipate that effective overlap may be reduced under field conditions and build in some margin to avoid failures. At sites where accuracy is unstable, instead of immediately blaming only the altitude, review whether your approach to overlap was realistic.

Overlap settings form the foundation that supports processing success rates and the stability of results. You may think you’re improving efficiency by reducing the number of photos, but it’s not uncommon to lose extra time to re-surveys or reprocessing. If the site presents even slight difficulty, it’s ultimately more efficient to choose overlap settings that prioritize connectivity and thereby strengthen the foundation of accuracy.

Setting 3: Fix the shutter speed and exposure conditions

In drone surveying, position information and flight routes tend to get most of the attention, but the quality of the photographs themselves also has a major impact on accuracy. Among these, the settings for shutter speed and exposure are particularly important. Because image processing aligns photos by detecting features within them, images with a lot of blur, blown-out highlights, or crushed blacks will produce unstable point correspondences. Even if a photo looks OK at a glance, if fine details are smudged or contrast appears unnatural, it can cause errors in surveying applications.

One particular point to watch is relying too heavily on automatic exposure during flight. At sites where ground surface colors are not uniform, or where sunlit and shaded areas coexist, brightness can vary greatly from photo to photo. Furthermore, when soil, grass, pavement, water surfaces, white materials, black weed-control mats, and the like are mixed, the camera will apply corrections each time, making it easy for the consistency of consecutive photos to break down. These variations can appear in post-processing as stitching errors or increased noise.

To improve accuracy, it is important to pay attention to a shutter speed that can suppress subject blur and aircraft movement blur. Even if a drone appears stationary, it is constantly subject to micro-vibrations and movement while flying—especially in wind. With a slow shutter speed setting, fine features on the ground will blur, making it difficult to stably pick up corresponding points during processing. Details that are important for accuracy checks—slope surfaces, areas covered with crushed stone, soil streaks, and areas around boundary stakes—are particularly susceptible to this kind of blur.

On the other hand, if you prioritize shutter speed too much, under dark conditions the sensitivity can increase excessively and image noise may rise. Photos with a lot of noise may look sharp at first glance, but in reality they tend to confuse feature-point detection and can be disadvantageous in terms of accuracy. Therefore, settings that improve accuracy are not simply the fastest ones; they suppress blur while maintaining the balance between brightness and noise.

Also, in surveying, consistency across consecutive photographs is more important than the visual beauty of color. If brightness or color tone varies greatly between photos, producing a seamless three-dimensional reconstruction becomes difficult. In particular, on days when clouds and sun alternate quickly, during the morning and evening when light comes in at an angle, and under surface conditions that create strong reflections, it is important to keep settings stable. Simply avoiding times of day when shooting conditions are likely to change can improve the stability of the results.

A common practical mistake when handling exposure settings is to arrive on site and fly using the default settings as they are. If you perform almost no test shots and try to get by with a single run, you may later notice blur or differences in brightness when reviewing the photos. In surveying, it is effective to take a few test shots before the flight to check how the ground appears and to look for blur, blown highlights, or crushed shadows. This extra step can significantly reduce the risk of needing retakes.

Furthermore, on windy days or at sites where ground-surface contrast is low, it is safer to manage photo quality more strictly than usual. In bare soil areas with little surface variation or on uniformly paved surfaces, even slight blur can make feature extraction difficult. Conversely, in areas with a lot of vegetation, the movement itself becomes a source of error, so instability in shooting conditions becomes a greater problem. In other words, it is not that you can be less careful with settings because the subject is simple; the more monotonous the surface, the more directly differences in shooting quality will affect accuracy.

Photos are the input data for the analysis itself. No matter how much you refine the flight path, if the original images are unstable the accuracy will not improve. Reviewing shutter speed and exposure settings is a very practical way to improve surveying accuracy. Precisely because these are elements that cannot be corrected after capture, they are among the settings you should finalize most carefully before flight.

Setting 4: Adjust flight speed and shooting intervals

The settings for flight speed and photo capture interval also have a major impact on the accuracy of drone surveying. On site, because teams want to shorten work time or reduce the number of battery swaps, they may be tempted to increase speed. However, if the speed is too high, the quality of each photo and the stability of image overlap tend to decrease. Even if photos are captured, increased ground motion and variability in shooting positions can cause distortions or gaps in the processing results.

There are two reasons why flight speed affects accuracy. One is that motion blur while moving becomes more likely. Even if you ensure a sufficiently fast shutter speed, subtle changes in aircraft attitude or wind-induced vibration make image stability more prone to degradation at higher speeds. The other is that the actual overlap can break down when combined with the capture interval. Even if you theoretically set a sufficient overlap rate, if speed and shooting timing are not synchronized, the required overlap may not be obtained.

Particularly in windy conditions, if flight speed is set carelessly, a drop in accuracy becomes pronounced. This is because the ground-relative speed changes between tailwind and headwind sections, causing the intervals between photos to become uneven. As a result, some sections may have excessive overlap while others suffer from insufficient overlap. Even if the overall processing succeeds, problems such as sudden elevation fluctuations in certain sections or disrupted seams are more likely to occur. These errors are troublesome because they are difficult to diagnose afterward.

Regarding the capture interval, shorter is not always better. If you make the interval too short, data volume balloons, processing load and management burden increase, and flight efficiency falls. Conversely, if the interval is too long, you won't achieve the required overlap and the stability of 3D reconstruction will be compromised. The important thing is to consider flight speed, altitude, the desired overlap rate, and how the ground is captured as a whole. Changing just one of these can cause inconsistencies with the other settings.

At sites with many structures or that include slopes, it is generally more stable to reduce speed. Around complex shapes, it is important to capture subtle changes in appearance accurately, and flying too fast can lead to missed information. On wide, flat areas you may be able to increase speed somewhat without issue, but for surveying purposes you should not be satisfied with merely capturing images—you must judge whether the results have the reproducibility necessary for comparison and for producing drawings.

In practice, increasing speed to shorten flight time can lead to the need for re-surveys or reprocessing, which increases overall man-hours. On site, attention tends to focus only on the duration of a single flight, but what truly needs shortening is the lead time for the entire survey. In that case, it is more rational to prioritize the stability of data capture and acquire data that are easy to process from the start. If high-accuracy results can be obtained in one pass, the whole process — including time for verification, correction, and explanation — will be faster.

Furthermore, flight speed also affects safety. In locations with nearby obstacles or at sites where wind direction is unstable, reducing speed improves the stability of the flight path. In surveying, following the planned line itself is important, so it is counterproductive to force higher speeds if that disrupts maintaining the planned path. The more you want to increase accuracy, the more you should choose settings that reliably capture consistent data rather than settings aimed at finishing quickly.

Flight speed and capture interval may appear to be efficiency settings on the surface, but are actually foundational conditions for accuracy. Aligning them with altitude, overlap rate, and image quality will greatly improve the stability of processing results. Rather than using the same values every time, adjusting them according to wind, terrain, the subject, and the required accuracy leads to more reproducible operations.

Setting 5 Review reference points and position correction settings

If you want to substantially improve the accuracy of drone surveying, it is essential to review not only image conditions and flight conditions but also how you establish positional references. By "positional references" I mean tying positions to known points on site, utilizing high‑precision position corrections, and ensuring consistency in how coordinates are handled. Even if photos are sharp, overlap is sufficient, and flight is stable, weak positional referencing will make it difficult to achieve the accuracy and reproducibility required for surveying.

One common problem in practice is that something can look relatively clean, yet not match when overlaid with existing drawings or past data. This happens when internal connectivity is consistent but alignment with external reference systems is insufficient. On sites where progress comparisons, as-built verification, earthwork volume differences, and checks against design data are performed, this discrepancy can become a major problem later on. The fact that a three-dimensional model appears coherent does not mean it can be used as survey deliverables.

What becomes important there is the placement of reference points and the configuration of position-correction settings. By establishing reliable positional references within the site, you can reduce distortion of the overall model and more easily improve horizontal and vertical stability. Especially on large sites or sites with significant topographic variation, if reference placement is biased, even if some areas have good accuracy, large deviations can occur in distant locations. It is important to plan reference placement while looking at the entire site — edges, central areas, and locations with elevation differences.

Even when using a position correction mechanism, if settings and operations are ambiguous you may not achieve the expected accuracy. For example, if the concept of coordinates differs between the acquisition side and the processing side, if the elevation reference within the site is not sufficiently understood, or if the comparison dataset has different reference conditions, it can become difficult to explain discrepancies in later processes. Improving accuracy is not simply making the numbers better; it means creating a state in which anyone can reuse the data under the same standards.

Whether to use reference points depends on site conditions and objectives, but in situations that require high reproducibility you should not overlook positional references. For example, if you only need a one-off general overview, a relative model may be sufficient, but for before-and-after construction comparisons, as-built verification, design checks, or overlaying multiple time points, the clarity of the reference points will determine the quality of the results. The more you genuinely want to improve accuracy, the more important it is not to try to solve everything with flight settings alone.

Additionally, on-site the visibility and placement condition of the control points themselves are important. Even if they are carefully installed, if they cannot be clearly identified in photographs the stability of processing will decrease. It is necessary to prepare conditions that make them easy to recognize—such as contrast with the surroundings, the flatness of the installation surface, how shadows fall, and the presence or absence of obstructions. Do not be satisfied with the mere fact that a control point has been placed; be mindful of whether it can be consistently confirmed within the captured data.

As the field person in charge, before the flight you should clarify "which data will be overlaid this time," "whether to prioritize horizontal accuracy or vertical accuracy," and "whether numerical justification will be required for comparisons or explanations," as this will make the approach to setting the reference clear. If you fly with these left ambiguous, problems will remain at the post-processing stage that cannot be remedied by changing settings. The final deciding factor in improving the accuracy of drone surveying lies not only in skillful shooting but in how you can make the positional information reliable.

On-site factors that commonly cause variations in accuracy apart from settings

So far we have introduced five settings to improve accuracy, but in practice, even when the settings are correct, results can still be unstable. This is because site conditions have a large impact on data acquisition. In other words, reviewing the settings is important, but that alone will not solve all accuracy problems. By understanding the site factors, you will be able to properly assess the significance of the settings.

First and foremost is the wind. Strong winds make the aircraft's attitude more prone to disturbance, causing flight-line deviation, image blur, and a decrease in effective overlap. Even if the settings are fine, wind can prevent the same conditions from being reproduced. Especially at sites where wind direction is not constant, images tend to differ between the outbound and return passes, resulting in differences in processing stability. If you want to improve accuracy, you should prioritize ensuring stable flying conditions before pushing the settings.

Next, the lighting conditions. Times when shaded and sunlit areas are heavily mixed, or when the sun is low and shadows are long, make surface features prone to vary from photo to photo. If a uniform appearance cannot be achieved, processing accuracy will be affected. Because simply shifting the shooting time slightly can sometimes stabilize the results, it is worth reviewing the shooting timing before tweaking the settings.

The condition of the ground surface is also important. Puddles, highly reflective surfaces, monotonous bare soil, swaying grass, movement of vehicles or workers, and changes in material placement—all of these become sources of error. For example, even when the surface looks the same, whether it is wet from a recent rain or dry can significantly change how it appears in photos. At sites where it is difficult to keep shooting conditions consistent, there are limits even if you carefully adjust settings. In such cases, you need to consider accuracy measures that also include the timing of photography and site tidying.

Also, at sites with many nearby structures or trees, the radio environment and visibility conditions cannot be ignored. Because the stability of positioning information and the freedom to safely capture the required area are reduced, operations identical to those used in flat, open terrain may not be feasible. At such sites, rather than searching for the correct settings, it is more realistic to plan—including how much to include in the scope of the current survey and which areas to supplement by other means.

In other words, improving accuracy cannot be completed solely through the onboard settings screen. It is important to read the site conditions, align the imaging conditions, organize the target area if necessary, and clarify the positional reference before optimizing the settings. If you get the order wrong, repeatedly fine‑tuning the settings alone will not produce stable results. All the more when you feel accuracy is lacking, you need to take a step back and review not only the configuration values but also whether the site itself was in a condition suitable for surveying.

Summary

Improving the accuracy of drone surveying isn't something you can achieve by doing just one special thing. The basic approach is to adjust five settings—capture altitude, image overlap, shutter speed and exposure, flight speed and capture interval, and the approach to ground control points and positional corrections—according to site conditions and project objectives. Each of these factors can be effective on its own, but in practice they are interrelated, and changing one setting affects the others. That's why, rather than reusing the same settings every time, it's important to decide them by working backward from the outcomes required for the current site.

In practice, people often feel relieved simply because the flight was completed and only notice lax settings when they later review the processing results. However, for surveying that requires accuracy, the flight is not the goal but the entry point. To acquire data usable in downstream processes in a single pass, decisions about settings before arriving on site are critically important. If you want to reduce re-surveys and reprocessing and operate with an eye toward volume checks, progress comparisons, and cross-checking with drawings, prioritize reviewing the five settings introduced here.

And to further leverage data acquired by drones on site, it is effective to consider not only aerial measurements but also ground-based position references and supplementary measurements. For example, in situations where you want to quickly carry out coordinate alignment checks, supplementary measurements at key locations, as-built verification, or overlays with photos and point clouds on site, a ground-side system capable of handling high-precision measurements is useful. If you want to make such operations smoother, combining an iPhone-mounted GNSS high-precision positioning device like LRTK can make it easier to expand the range of applications for results obtained from drone surveying. By maintaining a perspective that balances the accuracy of aerial acquisition with ease of handling on the ground, overall field surveying operations can be made even more efficient.