22 Apr MEASURING ACCURACY LEVELS FOR ORTHOPHOTO MAPPING WITH AND WITHOUT GCP!
At OFO Tech, we bring essential business information and industry knowledge to our clients. We believe in an open environment where stakeholders can engage in topics that matter. We invite you to our new blog, as we explore the latest topics and discussions on the commercial drone sector and related technologies of IR 4.0.
As a leading UAS Solution Provider, we have been asked on numerous occasions about Quality comparison and Accuracy from data using GCP. We have incorporated it as a standard SOP. OFO Tech delivers high level of precision to guaranteed customer satisfaction.
In this topic we will address GCP vs non-GCP and the differences in terms of accuracy levels of the actual Mapping results. We will discuss on Ground Control Points (GCP). Are they required in UAV mapping? Find out the differences!
OFO Tech is dedicated to offer best-in-market solutions to clients and will furnish you with everything you need to know. Follow us as we give expert opinion on UAV mapping in our informative blog on the advantages of using GCPs.
Our case study is open for everyone to participate with comments in the section below!
(*) GCP: Ground Control Point (GCPs are points which the surveyor can precisely pinpoint: with a number of known coordinates. The methodology is to enable to accurately map large areas).
The following will be a case study on the use of GCP for orthophoto mapping. Check out the differences as we methodologically present you the facts!
DATA ACQUISITION
Image Capturing
For our case example, please refer Table 1 to see the details planning for data acquisition. We have selected Wangsa Maju residential area as the site location.
For drones, we prefer DJI Phantom 4 Pro as our top pick, as it is ideal for any photogrammetric application and capable to withstand the toughest of duties.
The flying height is set at 88.20m above mean sea level. We followed the standard requirement for overlap and lateral overlap of 80% and 75% respectively. Like duh.
As we see in Table 2, shows the camera specification of DJI Phantom 4 Pro. This requirement is crucial in order to determine the quality of the image taken and to construct flight planning.
From Figure 1 shows, the hue from blue colour to red colour. The more overlapping images, the bluer shaded in map. Meanwhile, the red colour represents no overlapping images covered. The blue area has the highest precision and quality of the orthophoto produced.
The black dot represents the aerial photo taken. We succeeded to capture 876 images in coverage area. (Guess how many selfies we took).
GPS Data Collection
The real-time kinematic (RTK) method was used for GCP establishment. RTK positioning is a satellite navigation technique used to enhance the precision of position data derived from satellite-based positioning systems. Like yeah.
In Malaysia, the Malaysian Survey and Mapping Department (JUPEM) has produced MyRTKnet; the infrastructure that has been formed by the Global Navigation Satellite System (GNSS) reference stations and control centre to provide the GNSS data correction in order to give the position in real time.
From Table 3, we have collected a total of 9 points scattered around the coverage area. It contained the planimetric coordinate of XY values and ellipsoidal height (H); represents as Z value.
DATA PROCESSING
GCP ESTABLISHMENT
Continuing with data processing, we use two different processing methods in order to show the comparison between orthophoto with and without GCPs.
From Table 4, shows the flight information of survey data and the same image data were used to produce orthophoto with GCP and without GCP.
Figure 2 shows the location of GCPs in coverage area, Wangsa Maju.
Orthophoto with GCP
For processing with GCP, we used 5 points of GCP and 4 points for measurement checking and accuracy verification, we called as checkpoint (CP). We can see the GCPs and CPs located in Figure 2.
Based on Table 5, these are the coordinates values produced with GCP as reference during processing.
Orthophoto without GCP
For processing without GCP, we disregarded the RTK values as reference. The orthophoto was only produced by using the image coordinates. Every image taken by drone contains its coordinates. We established 9 points at the location stated in map Figure 2 for comparison analysis.
From Table 6, these are the processed coordinates values produced based on the geotag information of the aerial photo only.
ANALYSIS AND RESULTS
COMPARISON BETWEEN ORTHOPHOTO WITH AND WITHOUT GCP
The comparison of both produced orthophotos were evaluated with GCP processed orthophoto and RTK coordinates. Then, calculated Root Mean Square Error (RMSE) is used to evaluate 3D coordinates comparison.
From Table 7, the Northing and Easting values shows no difference compared to RTK coordinates. However, the Height values of PT7-CP shows the highest difference of 0.056m and PT9-CP is the lowest difference with -0.109m. The RMSE values for Northing and Easting are 0.000m, followed by Height with 0.044m.
As can be seen Table 8 shows the highest difference value of Northing is PT1 with -1.815m and lowest is PT9 with -1.220m. Then, the highest difference value of Easting is PT2 with 2.195m and the lowest is PT3 with 1.443m. Whereas, the highest difference values of ellipsoidal height is PT2 with 3.673m and lowest is PT4 with difference values of -0.212m. All the RMSE difference values between orthophoto without GCP and RTK in Northing, Easting and height are 1.4224m, 1.6846m, and 2.2919m respectively.
Figure 3 until Figure 11 show, the result of comparison between orthophoto with and without GCPs. All of the images are the processed images of orthophoto. The ‘×’ symbol represents the point of orthophoto without GCP while the circle symbol represents the actual GCP points processed with RTK. As we can see there, the difference distance of each point is in the range of 1.10m to 2.01m apart.
From Table 10, shows that RMSE values of orthophoto with GCP is superior than orthophoto processed without GCP. Hence, the results proven beyond doubts that the accuracy and precision of orthophoto mapping is greater with GCP.
CONCLUSION
As an enthusiast of surveying and mapping professionals, any successful projects have to come with accuracy; and that is the key of it. In a nutshell, this case study showing that the map accuracy can be highly improved from more than 2m to below 10cm for each axis of Northing, Easting and Height.
Using Drones can produce spectacular images. Nevertheless, when it comes to drone precision mapping it is the methodology and devices used that are the most important element required to deliver precision and accuracy
As our two cents, RTK data is very important to keep the consistency quality of the map. Of course, we do not claim that orthophoto mapping without GCP cannot be used for any application. Nevertheless, in order to achieve high quality, we have to go with additional processes for checking the accuracy and precision for orthophoto mapping.
Consequently, in our services at OFO Tech, we strive to offer the best of the best to ensure satisfaction for customers we connect with in the business landscape. We go through triple constraints to achieve good quality, fast product delivery and affordable cost.
Our conclusion dictates, that processing with GCP is better than depending purely on drone coordinates; as deployment of GCP meets the golden triangle of project management for all orthophoto mapping project onwards.
For Surveying, Architects, Project Planners and other mapping professionals requires precision mappings and highly accurate maps. We at OFO TECH always delivers cost effective, on time Precision and Accurate Maps to meet industry and professional demands.
At OFO Tech we have plenty more tools and techniques to improve your business bottom line. We are eager to answer all your burning questions and satiate your curiosity. This study case was just one of them. More sharing discussion and article coming up next.
OFO TECH has the technology, know-how, methodology and tools to achieve our Quality Goals commitments, Comprehensive and Integrated Solutions.
OFO TECH will always strive to deliver Quality, Accuracy and Precision.
Can you keep a secret? Because we can’t!
We are keen to share more top secret knowledge and expertise with our reader. Knowledge is power after all. Subscribe to our mailing list to be updated on all latest blog updates, insight and knowledge sharing on our website!
Sr Dr Zainal Bin A Majeed
Posted at 03:24h, 05 JuneI must say this research is fantastic. I had done quite similar kind of research before which I recommend to JUPEM that drone mapping is useful for mapping topographical medium scale map. Cost for aerial photography activities considering flying normal aero-plane is much too expensive.
Keep it up. Congrats!
admin
Posted at 04:19h, 10 JuneWe greatly appreciate your comment. Indeed by using drone mapping is much more cost effective than conventional approach. We hope to hear more from you, thank you for subscribe!
Our upcoming blog post will be update soon 😄