Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.
We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.
Remotely Piloted Aircraft (RPAs), commonly called drones, have established themselves as a valid remote sensing platform. These platforms, capable of flying on demand and in often otherwise inaccessible environments, have proven themselves to serve a niche data product where both high spatial and temporal scales are needed by the researcher. The use of RPAs as a research tool often comes the need to establish a high degree of horizontal and vertical locational precision outside of what conventional GNSS provides. Ground Control has been traditionally established with the use of ground surveyed Ground Control markers but is increasingly being established using Real Time Kinematic and Post-Processing Differential Correction methods. In forestry and natural resources science, being able to employ a reasonable degree of precision and accuracy is essential in the utilization of RPA as a data collection tool, yet past and present research trends show that no one method is superior to another and that different applications call for different forms of correction when factors such as cost, time, and efficiency are applied. This chapter explores the use of ground control in RPA data acquisition.
Purdue University School Aviation, West Lafayette, IN, USA
Aishwarya Chandraskaran
Purdue University Department of Forestry and Natural Resources, West Lafayette, IN, USA
*Address all correspondence to: jhupy@purdue.edu
1. Introduction
Forests, in their many forms, often cover vast expanses of territory and can prove difficult to traverse for data collection in research and inventory practices. Remotely Sensed Imagery has long been used within the forestry community to study and manage forest communities in a wide array of temporal and spatial scales [1]. Foresters were amongst the first to recognize the value of aerial imagery coming first from fixed wing aircraft, then satellites, and now Remotely Piloted Aircraft (RPAs) [2]. What are commonly referred to as drones, RPAs fill a valuable niche as a remote sensing tool for the forest researcher for their nimble deployment abilities and gathering imagery with ground spaced distance resolutions of 1–2 cm [3]. For researchers studying forest environments, especially those studying a dynamic forestry community such as scrubland where changes can occur in rapidly over time, having the means to remotely sense these communities on a temporal resolution of choice can prove invaluable [4, 5].
Just like any other remote sensing platform, RPAs have benefits and drawbacks, and it is up to the forest researcher to decide the best form of remote sensing technology to be utilized. The quality RPA imagery, like any other form of remote sensing data, depends on whether the right methods and practices were implemented to gather that data. Imaging forest environments with RPA sensor platform can present many challenges to the forest researcher, especially in forested environments with dense forest canopy [6]. That stated, the low woody plants interlinking low grasslands and dense forest communities afforded by scrub forest provides the researcher with ample means to utilize RPAs as a valid Remote Sensing tool.
Although Scrub Forest communities may not be nearly as challenging to the researcher as those trying to effectively image the dense canopies of mature temperate forest, there are still challenges that should be taken into account for quality image acquisition [7]. Namely, these are: (1) Shadows and shade generated by the vegetation; (2) Wind movement on vegetation leaves and branches; and (3) achieving spatial accuracy and precision with effective ground-based control measurements (Figure 1). Engaging in data collection at the right time of day, and with the right weather conditions can alleviate the first two possible sources of error, but gathering proper ground control requires specific forms of hardware, knowledge, and methods for effective ground control measurements [8]. Implementation of Ground Control is of special importance in research with a temporal component, or when vertical measurements matter such as those measuring scrub forest height and canopy measures. In scrub forest studies where trees can rapidly achieve heights of 4–5 m over several years, having proper vertical data measures with ground control is crucial for the researcher [9, 10, 11, 12].
Figure 1.
RPA aerial survey sources of error examples.
In this chapter we provide an overview of the hardware available to the forest researcher looking to instate proper ground control in their RPA data acquisition efforts. These same technologies can often also be applied to those looking to gather other forms of ground truthing data for ground-based data collection efforts. We will cover the advantages and advantages of the various forms of technology, mainly by balancing quality with cost and the expertise needed. For example, in some cases where temporal data and vertical measures are not needed, there may not be a need for expensive hardware to achieve the results needed. We will begin with an overview on how ground-based measurements have traditionally been performed, and then provide an overview of Global Navigation System Satellite (GNSS) technology and how it relates to ground based measurements, especially those involved with achieving geolocational control.
There are two basic ways one can gather locations of features on the earth’s surface. One is the use of remotely sensed data, and the other is to rely on field-based measurements. Traditional forest survey methods relied heavily on a compass, clinometer, and distance measuring devices to survey trees and survey boundaries [13, 14]. For example, for a forester to survey what are pioneer, established, and relic species of a forest environment, they would use a point-quarter survey method. In this method, a series of random points are set throughout the forest and the distance to the nearest tree above and below 5 cm diameter at breast height are surveyed in four quadrants around the point. This method relied on a compass, distance measure, tree diameter measure, and sometimes a height clinometer. Results from surveys such as these relied on extrapolative statistics in terms of stating forest composition, and the location of each point quarter survey location were often based implicitly located based on reading a hard copy map or imprecise GPS locations [14]. Today, much of the survey work conducted relies on Global Navigation Satellite System (GNSS) technology in conjunction with some form of remotely sensed imagery [15]. The higher the resolution of the remotely sensed imagery, the more one can do in terms of engaging in forest surveys, yet the quality of the GNSS data also relates to how much precision and accuracy is provided with the ground survey data in terms of supplementing the remotely sensed imagery.
2.1 GNSS technology
Global Positioning System (GPS) is often used synonymously with Global Navigation Satellite System (GNSS) [16, 17]. Both technologies rely on a receiver station that utilizes signals from a constellation of satellites providing positional and timing data from space, yet GPS satellites are those associated with the United States, while GNSS is comprised of navigation satellites fielded by a wide array of participating countries (Table 1). GNSS receivers use this data to derive location information of a point. The number of satellites available far exceeds those available using only GPS satellites, to where the number of satellites available is often more than 14. The higher the number of satellites, the less dilution of precision error (DOP) levels one has, and the more precise their location [18]. GNSS technology is not without error, with one of the more common errors that those using GNSS in forested environments experience is multi-path error to where the signals being picked up by the receiver get sent in multiple directions when they hit leaves and other objects above [18]. Sometimes this error makes ground-based survey with conventional GPS receiver hardware near impossible. Fortunately, there are ways to lessen or overcome many of the sources of error by using dual frequency receiver technology, or by incorporating the use of a base station and GPS rover in the survey. These methods and the associated technology are covered in more detail in a later section.
Global navigation satellite system network participating nation states and satellite information.
2.2 Errors in GNSS
When utilizing GNSS technology, it is important to understand some sources of error propagation which might affect the overall precision and accuracy of the location measurements [19, 20, 21, 22, 23]. Atomic clocks present within the satellite can be slightly inaccurate due to drift- a phenomenon where the clock does not run at the same rate as a reference clock. GNSS uses multiple satellites for establishing a location on ground and a drift between these clocks can introduce positional errors. For example, 20 ns of clock drift can result in a ± 6 m of positional error at the receiver end. This type of error can be corrected by looking at the downlink data for clock offset and offer corrections. Another significant impact on positional accuracy can be due to orbital errors. GNSS satellites travel precisely in a consistent orbit and any variation within the orbit can have significant positional errors. These orbits are continually monitored by the GNSS ground control system and corrections are sent within minutes of noticing any change in the orbit. Even with the corrections, we can expect positional errors up to ±2.5 m. The next source of error results from the ionic activity at the ionosphere. The ionosphere is a layer of the atmosphere from 50 to 1000 km above the earth containing electrically charged particles. These particles can alter the satellite signals causing huge positional errors around ±5 m. This type of error is known as ionospheric delay and can depend upon the amount of solar activity, time of day, season. Year, and location. GNSS receivers receiving multi-frequency signal for example L1 and L2 (frequency bandwidth) can use this measurement for comparing any differences/delays in the signal and offer correction on the position. Similar to ionospheric delay, tropospheric delay can also introduce positional errors. Although less likely, these errors are caused due to changes in humidity, temperature and pressure within the atmosphere. This type of error is local and will be experienced by the base station and rovers in a similar way. Utilizing tropospheric models, Differential GNSS or real time systems can reduce this error. There are errors associated with the receiver known as the receiver noise which is inevitable and constitutes about 1% of the wavelength of the signal involved. It is a relatively small contributor of position error and can be minimized by using high end receiver stations. The final but most common type of positional error, especially working with forested environments referred to as multi-path error. This error occurs when the receiver receives the satellite signal from multiple paths. This error can be reduced by moving to a location unobstructed by tree canopy which is nearly impossible. Connecting to multiple satellites (10–12) can improve positional precision and accuracy as well [21, 22, 23].
2.3 Establishing ground control
Many forest researchers may have implemented some form of GNSS technology in their research to perhaps locate study area boundaries, or to ground truth some patterns or vegetation noted in remotely sensed imagery, be it from satellite, traditional aircraft, or even RPA data (Figure 2). It is important to note here that ground truthing is a form of confirming on the ground what was noted in remotely sensed imagery, and often that can be done with GNSS technology that does not require a large degree of precision in terms of coordinate locations [24, 25]. Ground control, however, differs from ground truthing in that the coordinate locations gathered by that GNSS hardware should be better than that of the RPA platform as they serve to correct both the horizontal and the vertical location of the imagery gathered [26]. Control points are points identified and established on the ground whose location is known with sub meter to sub centimeter accuracy. These points and their coordinate measures are used for correcting and establishing coordinates for the final map [7].
Figure 2.
RPA student researchers recording GCP location.
There are different ways to extract the coordinates of known point locations on the ground. In certain situations when the user has existing data that is locationally precise and accurate, one can obtain coordinates from easily recognized locations and enter those values for control [27]. In urban and developed environments, this is easy to do with readily available high-quality data, and plenty of anthropogenic features to choose from [28]. For example, if a user wanted to apply control to RPA imagery in a given suburban neighborhood, they could look up existing control points, or perhaps use existing corrected imagery to find coordinates over sidewalk intersections or storm drain covers. For the forest researcher, however, quality datasets with readily recognized features can be hard to come across and therefore establishing ground control will require corrections that relies upon the precision and accuracy afforded by GNSS technology in three primary methods: 1) deployment of ground-based survey markers; 2) Post-Processing Kinematic corrections 3) Real-Time Kinematic Corrections (Figure 3).
Figure 3.
Illustration of different control used in surveying.
2.3.1 Survey markers
Consumer grade GPS traditionally were based upon single frequency receiver technology, but are increasingly being replaced with dual frequency hardware, and make use of the GNSS network [29]. While dual frequency eliminates a great deal of multi-path error, most consumer GPS have no means to apply corrections to the data, and therefore the data has quite of bit of variability in terms of precision and accuracy [18]. Because this data is gathered on the ground, the quality of the data is often of a worse quality than the data gathered for the RPA platform and introducing this form of data results in a diminished quality dataset. Consumer grade GNSS units should be avoided in any type of ground control survey [30].
If the aerial survey being conducted has openings with a clear view to the open sky, traditional ground control markers, with a survey grade GPS, are possible. In many forest studies, the use of traditional ground control is negated because the Ground Control Point (GCP) markers are obscured by forest canopy, along with the signal to record those markers being diminished by multi-path error in the leaves of the canopy [31]. Scrub-forest studies, however, allow ground control with traditional markers if they are placed in locations where the signal is not obscured by tall objects such as canyon walls or tree canopies. Previous research has demonstrated that even if a GCP marker is placed near a building or tree, the Root Mean Square (RMS) error of the point is significantly impacted [32].
GCP markers come in a variety of forms, ranging from merely painting a ‘V’ shape mark on the ground, to temporary markers that can easily be carried and deployed, to those that are anchored in place for long-term studies [7]. The most important aspect of any GCP marker is that it can be seen in an aerial image and that the survey point’s location is known. Most ground control is recorded at the center of the GCP marker, and this is why most are marked in a manner to where the center of that point is located. Sometimes, however, such as in situations where temporary markers are painted in, it is important that those engaged in the ground survey indicate where the survey point was recorded, as even being centimeters off, can drastically diminish the quality of survey results [33]. Placement of GCPs in any survey should strive toward even placement of the markers across the study area. While field conditions often make ideal GCP placement impossible, the researcher should take care not to cluster markers or only place GCPS along the study area’s perimeter.
In selecting a GNSS hardware device to record the location of GCPs, great care should be exercised regarding how the surveyed points will be corrected [34]. Many users assume that just because they have a dual frequency survey-grade GPS, then they need not worry about precision and accuracy. The user needs to be aware that corrections still need to be applied. The good news for the forest researcher is that the cost of survey grade hardware has come down tremendously in recent years. As little as 5 years ago, the cost of a survey grade GNSS unit would likely exceed $20,000 USD, to where now one can purchase a reasonable hardware configuration for under $2000 USD. This cost is still much higher than a consumer grade GPS, but well worth the investment when location is needed for temporal studies, or those that require establishing proper vertical surfaces [35]. A typical survey GPS configuration consists of a rover receiver connected to a base station. In the United States, the Continuously Operating Reference System (CORS) Network can be used to apply corrections either in real time if connected to a cellular network, or through post-processing of the rover receiver data [36] (Figure 4). It should be noted that the quality of the data diminishes the further one gets from a CORS base station to where after 5 cm, the error is about 1 cm/km. In most locations within the United States, this is not an issue as the network is very dense and often allows the user to pick a base station within close range. In certain situations, though, where such CORS base station not available, or where a cellular network does not allow for real time corrections, the user may need to set up their own base station and establish a connection with their rover receiver using a radio connection [33]. In this case, the base station is set up over a known point such as a survey benchmark, or a pre-established fixed point. Here a cautionary note needs to be applied in that the coordinates provided for the base station are relative to the rover and if the base station location is off, so will all the data. Users wishing to establish their own base-station control points will need to record data in static mode at that location for a minimum of 2 h, and then get that data corrected through online services such as OPUS (United States) or ONPOZ (Most International) before using those coordinate values [37]. Many researchers prefer using their own established base station and rover to gather control as the time to gather a point can be quicker, and there is less risk of losing cellular connection to establish a fixed GNSS location to record a point location [38]. This does require more investment in hardware as the user would have to purchase two GNSS hardware devices though, one as the rover, and one as the base station. The base station also requires some type of a leveling tripod for setup, and most rover units rely on some type of a mobile tripod for stability in recording a point.
If the thought of placing and recording GCP markers before every survey seems daunting, there is an option to place GCP markers that have static GNSSS recording devices embedded within them [7, 39]. One example would be Propeller Aeropoint markers that were developed for the Australian mining industry as it was recognized that GCP markers in the extraction industry take a quite a long time to place and record and are very difficult to make permanent. Each Aeropoint marker has its own GNSS receiver, battery, and solar panel to ‘trickle’ charge the battery during operation [40]. After the unit is placed on the ground, the power is turned on and the Aeropoint marker immediately begins static data collection. When the aerial survey is complete, the marker is turned off before being picked up [35]. Data from the devices can then be wirelessly uploaded to cloud-based post-correction software and corrected using either the CORS network or a user established base station [41]. While the ease of utilizing these markers is notable, it does come with a fairly high hardware investment and an annual subscription fee.
2.3.2 Platform-based PPK and RTK
Another efficient way of observing location information without placing any ground markers is by utilizing Real Time Kinematic (RTK) and Post Processing Kinematic (PPK) technology (Figure 5). Both these technologies have a similar operation where each image captured during a survey is geotagged through an onboard GNSS recording unit [33, 42]. For both these systems, the GNSS unit triggers image capture during flight. This can be set based on the desired image overlap/distance or through time. Both these systems work on the principle of differential positioning – that two receivers that are close by will experience similar positioning errors. With RPA mapping, a reference base station with a known location acts as one of the receivers while the drone acts as the other. As the location of the base station is known with high precision, it can calculate the absolute position of the satellites based on its position and signal navigation time. These corrections can either be transmitted directly to the drone while mapping as in RTK technology or downloaded and corrected postflight as in PPK [43].
Figure 5.
Location measurement procedure GCP vs. PPK vs. RTK.
The reference station for using RTK technology can either be your own established base station established on-site or utilize some of the established base stations from the National Oceanic and Atmospheric Association Continually Operating Reference Station (NOAA CORS) network. In the case of utilizing a CORS station, the measurement results can be directly transmitted to the drone in real-time via a communication channel. When applying these corrections, we can expect sub-meter to sub-centimeter location accuracy. However, it is important to keep in mind, for most forest mapping, the location is remote with valleys, ridges, mountains, and trees that can interrupt communication with the base station. In such cases, where a successful connection cannot be established between the reference base station and the drone, positional corrections cannot be made leading to failed missions [44]. Another factor to consider is that the RTK corrections are made relative to the base station, so if the coordinates of the base station are not accurate, the corrections made are relative to the base station coordinates. A final factor to consider is that an RTK system is very expensive and there are no means of correcting positional errors post flight [33, 45].
However, when using PPK technology, every image location is collected and stored during flight using the on-board GNSS unit [33]. This unit can either be purchased as a PPK kit which contains a GNSS receiver along with the platform and sensor or can be assembled in-house based on the platform type. This is to ensure seamless installation and operation. The cost for a PPK unit can range anywhere from $500–$10,000. Post flight, either using an established base station or through the CORS network, we can retrieve the accurate location differences and perform carrier phase correction for the RPA image locations. This way, we are eliminating the issue of data connectivity between the reference base station and the drone during flight and can access the base station location information long after data collection mission [45]. There are numerous post processing software available to perform geolocation correction for the RPA imagery. One such software is EZ Surv where users can input antenna and camera position offsets (from the GNSS unit), location log file collected by the onboard GNSSS unit, CORS station information, and the coordinate system in which the image locations were recorded to perform spatial corrections and produce sub-centimeter image location precision [46]. In this case of post processing, it is important to make sure that the total number of images collected, and the images registered within the location log file match. Any mismatch between these can result in a spatially erroneous orthomap. Also, the postprocessing software has a cost associated either with a single purchase or as a membership with continued renewals.
For the scrub forest researcher, Remotely Piloted Aircraft (RPAs) can be a valuable remote sensing tool, capable of gathering imagery at both high spatial and temporal resolutions. They are also nimble in nature, with the capability of being launched at the study site at a near on demand basis. Using RPAS for remote sensing purposes in scrub forest environments should also come with an awareness of how best to achieve the best precision and accuracy available. While certain methods should be applied in capturing imagery that relates to taking clear imagery with minimal shadowing, blurring, and distortion, proper use of ground control may be arguably one of the most important factors to consider. Ground is essential for temporal studies that look at transitions over time, such as ecotonal shifts in scrub forest boundaries. This is because the imagery must overlap in temporal analysis and thus precision and accuracy must be in place for this to occur. Likewise, if the researcher wishes to engage in any type of analysis where vertical measurement takes place, then implementation of proper ground control is a must. Researchers should avoid using uncorrectable GNSS hardware as the use of that hardware will often diminish the quality of the RPA imagery. Ground control can be achieved through placement of markers on the ground that would show up in an aerial survey. Scrub forest environments with ample openings to the sky make this option available, but the researcher should be aware that placement of GCPs needs to occur across the entire study area, and sometimes conducting such surveys can be timely, especially when dealing with large study areas. Post-Processing Kinematic (PPK) and Real-Time Kinematic corrections allow for the researcher to make corrections using a base station but are present larger investments in RPA platform and hardware technologies. The good news for the scrub forest researcher is that these technologies now offer the means to engage in more robust, precise, and accurate surveys over scrub forest than ever before.
References
1.Markham K, Frazier AE, Singh KK, Madden M. A review of methods for scaling remotely sensed data for spatial pattern analysis. Landscape Ecology. 2023;38(3):619-635
2.Tang L, Shao G. Drone remote sensing for forestry research and practices. Journal of Forest Research. 2015;26:791-797. DOI: 10.1007/s11676-015-0088-y
3.Guimarães N, Pádua L, Marques P, Silva N, Peres E, Sousa JJ. Forestry remote sensing from unmanned aerial vehicles: A review focusing on the data, processing and potentialities. Remote Sensing. 2020;12(6):1046
4.Hartley RJ et al. A mixed methods approach for fuel characterisation in gorse (Ulex europaeus L.) scrub from high-density UAV laser scanning point clouds and semantic segmentation of UAV imagery. Remote Sensing. 2022;14(19):4775
5.Suo C, McGovern E, Gilmer A. Coastal dune vegetation mapping using a multispectral sensor mounted on an UAS. Remote Sensing. 2019;11(15):1814
6.Fraser BT, Congalton RG. Issues in unmanned aerial systems (UAS) data collection of complex forest environments. Remote Sensing. 2018;10(6):Art. no. 6. DOI: 10.3390/rs10060908
7.Cromwell C, Giampaolo J, Hupy J, Miller Z, Chandrasekaran A. A systematic review of best practices for UAS data collection in forestry-related applications. Forests. 2021;12(7):Art. no. 7. DOI: 10.3390/f12070957
8.Harwin S, Lucieer A. Assessing the accuracy of georeferenced point clouds produced via multi-view stereopsis from unmanned aerial vehicle (UAV) imagery. Remote Sensing. 2012;4(6):Art. no. 6. DOI: 10.3390/rs4061573
9.Siewert MB, Olofsson J. Scale-dependency of Arctic ecosystem properties revealed by UAV. Environmental Research Letters. 2020;15(9):094030
10.Wang H et al. Landscape-level vegetation classification and fractional woody and herbaceous vegetation cover estimation over the dryland ecosystems by unmanned aerial vehicle platform. Agricultural and Forest Meteorology. 2019;278:107665
11.Fraser RH, Pouliot D, Van Der Sluijs J. UAV and high resolution satellite mapping of forage lichen (Cladonia spp.) in a rocky Canadian shield landscape. Canadian Journal of Remote Sensing. 2022;48(1):5-18. DOI: 10.1080/07038992.2021.1908118
12.Assmann JJ, Kerby JT, Cunliffe AM, Myers-Smith IH. Vegetation monitoring using multispectral sensors – Best practices and lessons learned from high latitudes. Journal of Unmanned Vehicle System. 2019;7(1):54-75. DOI: 10.1139/juvs-2018-0018
13.Kiser J, Solmie D, Kellogg L, Wing MG. Efficiencies of traditional and digital measurement technologies for forest operations. Western Journal of Applied Forestry. 2005;20(2):138-143
14.Johnson CE, Barton CC. Where in the world are my field plots? Using GPS effectively in environmental field studies. Frontiers in Ecology and the Environment. 2004;2(9):475-482. DOI: 10.1890/1540-9295(2004)002[0475:WITWAM]2.0.CO;2
15.Goodbody TRH, Coops NC, Marshall PL, Tompalski P, Crawford P. Unmanned aerial systems for precision forest inventory purposes: A review and case study. The Forestry Chronicle. 2017;93(01):71-81. DOI: 10.5558/tfc2017-012
16.Blewitt G. Basics of the GPS technique: Observation equations. Geodynamics Applied GPS. 1997;1:46
17.Leick A, Rapoport L, Tatarnikov D. GNSS Positioning Approaches. In: Leick A, Rapoport L, Tatarnikov D, editors. GPS Satellite Surveying. 2015. DOI: 10.1002/9781119018612.ch6
18.Li X et al. Accuracy and reliability of multi-GNSS real-time precise positioning: GPS, GLONASS, BeiDou, and Galileo. Journal of Geodesy. 2015;89(6):607-635
19.Abidin HZ. 8 fundamentals of GPS signals and data. Manual of Geospatial Science and Technology. CRC Press; 2010. pp. 115-133. DOI: 10.1201/9781420087345
21.Håkansson M, Jensen ABO, Horemuz M, Hedling G. Review of code and phase biases in multi-GNSS positioning. GPS Solution. 2017;21(3):849-860. DOI: 10.1007/s10291-016-0572-7
22.Karaim M, Elsheikh M, Noureldin A. GNSS error sources. Multifunctional Operation and Application of GPS. InTech; 30 May 2018. DOI: 10.5772/intechopen.75493
23.Joardar S, Siddique TA, Alam S, Hossam-E-Haider M. Analyses of different types of errors for better precision in GNSS, 2016 3rd International Conference on Electrical Engineering and Information Communication Technology (ICEEICT). Dhaka, Bangladesh; 2016. pp. 1-6. DOI: 10.1109/CEEICT.2016.7873127
24.Keefe RF et al. Positioning methods and the use of location and activity data in forests. Forests. 2019;10(5):458
25.Yu K, Rizos C, Burrage D, Dempster AG, Zhang K, Markgraf M. An overview of GNSS remote sensing. EURASIP Journal on Advances in Signal Processing. 2014;2014(1):134. DOI: 10.1186/1687-6180-2014-134
26.Miller Z, Hupy J, Hubbard S, Shao G. Precise quantification of land cover before and after planned disturbance events with UAS-derived imagery. Drones. 2022;6(2):Art. no. 2. DOI: 10.3390/drones6020052
27.Iglhaut J, Cabo C, Puliti S, Piermattei L, O’Connor J, Rosette J. Structure from Motion photogrammetry in forestry: A review. Current Forestry Reports. 2019;5(3):155-168. DOI: 10.1007/s40725-019-00094-3
28.Martínez-Carricondo P, Agüera- Vega F, Carvajal-Ramírez F, Mesas-Carrascosa F-J, García-Ferrer A, Pérez-Porras F-J. Assessment of UAV-photogrammetric mapping accuracy based on variation of ground control points. International Journal of Applied Earth and Observation Geoinformation. 2018;72:1-10
29.He H, Li J, Yang Y, Xu J, Guo H, Wang A. Performance assessment of single-and dual-frequency BeiDou/GPS single-epoch kinematic positioning. GPS Solution. 2014;18:393-403
30.Deckert C, Bolstad PV. Forest canopy, terrain, and distance effects on global positioning system point accuracy. Photogrammetric Engineering and Remote Sensing. 1996;62(3):317-321
31.Odolinski R, Teunissen PJ. Single-frequency, dual-GNSS versus dual-frequency, single-GNSS: A low-cost and high-grade receivers GPS-BDS RTK analysis. Journal of Geodesy. 2016;90(11):1255-1278
32.Naesset E, Jonmeister T. Assessing point accuracy of DGPS under forest canopy before data acquisition, in the field and after Postprocessing. Scandinavian Journal of Forest Research. 2002;17(4):351-358. DOI: 10.1080/02827580260138099
33.Miller ZM, Hupy J, Chandrasekaran A, Shao G, Fei S. Application of postprocessing kinematic methods with UAS remote sensing in forest ecosystems. Journal of Forestry. 2021;119(5):454-466
34.Bolkas D. Assessment of GCP number and separation distance for small UAS surveys with and without GNSS-PPK positioning. Journal of Surveying Engineering. 2019;145(3):04019007. DOI: 10.1061/(ASCE)SU.1943-5428.0000283
35.Villanueva JKS, Blanco AC. Optimization of ground control point (GCP) configuration for unmanned aerial vehicle (UAV) survey using structure from motion (SFM). The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. 2019;XLII-4-W12:167-174. DOI: 10.5194/isprs-archives-XLII-4-W12-167-2019
36.Snay R, Soler T. Continuously operating Reference Station (CORS): History, applications, and future enhancements. Journal of Surveying Engineering. 2008;134:95-104. DOI: 10.1061/(ASCE)0733-9453(2008)134:4(95)
37.Han J-Y, Wu Y, Liu R-Y. Determining the optimal site location of GNSS base stations. Boletim de Ciências Geodésicas. 2012;18(1):154-169. DOI: 10.1590/S1982-21702012000100009
38.Hill AC, Limp F, Casana J, Laugier EJ, Williamson M. A new era in spatial data recording: Low-cost GNSS. Advances in Archaeological Practice. 2019;7(2):169-177. DOI: 10.1017/aap.2018.50
39.Tmušić G et al. Current practices in UAS-based environmental monitoring. Remote Sensing. 2020;12(6):Art. no. 6. DOI: 10.3390/rs12061001
40.Ucgun H, Yuzgec U, Bayilmis C. A review on applications of rotary-wing unmanned aerial vehicle charging stations. International Journal of Advanced Robotic Systems. 2021;18(3):17298814211015863. DOI: 10.1177/17298814211015863
41.Wang J, Ge Y, Heuvelink GBM, Zhou C, Brus D. Effect of the sampling design of ground control points on the geometric correction of remotely sensed imagery. International Journal of Applied Earth and Observation Geoinformation. 2012;18:91-100. DOI: 10.1016/j.jag.2012.01.001
42.Tomaštík J, Mokroš M, Surový P, Grznárová A, Merganič J. UAV RTK/PPK method—An optimal solution for mapping inaccessible forested areas? Remote Sensing. 2019;11(6):Art. no. 6. DOI: 10.3390/rs11060721
43.Puente I, González-Jorge H, Martínez-Sánchez J, Arias P. Review of mobile mapping and surveying technologies. Measurement. 2013;46(7):2127-2145. DOI: 10.1016/j.measurement.2013.03.006
44.Bakuła M, Oszczak S, Pelc- Mieczkowska R. Performance of RTK positioning in forest conditions: Case study. Journal of Surveying Engineering. 2009;135(3):125-130. DOI: 10.1061/(ASCE)0733-9453(2009)135:3(125)
45.D. Dinkov and A. Kitev, Advantages, Disadvantages and Applicability of GNSS Post-Processing Kinematic (PPK) Method for Direct Georeferencing of UAV Images. 2020
46.Effigis. EZSurv User Manual. Montreal, Canada: Effigis; 2019
Written By
Joseph P. Hupy and Aishwarya Chandraskaran
Submitted: 26 February 2024Reviewed: 04 March 2024Published: 02 April 2024
Open access peer-reviewed chapter - ONLINE FIRST
