High accuracy using drones, in a 3500ha Urban Cadastre, Francisco de Orellana 2017
Actualizado: 16 de ago de 2019
The following case study shows how with good flight planning and GCP distribution you can derivate products inside the tolerance for 1: 1000, with drones.
In the province of Francisco de Orellana, a survey of 3500ha was carried out for an urban cadastre, at a scale of 1: 1000. The expected tolerance on the scale is 0.3mm * 1000 (scale) = 300mm - 30cm.
It is essential to do proper flight planning and the distribution of homogeneous control points on the ground, to ensure that then the union of all the flights will be formed as a single photogrammetric block, an overlap between 20% blocks was established, also the control points were placed in their overlap.
Next, ideal flights were designed with the following characteristics:
Autonomous Vehicle Used WINGQUAD
Flight height: 150m
Sensor: Sony A5100 24Mpx
Speed: 16m / s
Duration: 1 hour
Covered area: 200ha
Next, the network of photogrammetric support points (GCPs) were designed with the following characteristics:
Receiver: GPS L1 / L2
Measurement type: Differential
Method: Rapid - Stactic
Session time: 30min
Short Baseline: <15km
The adjustment is made with the control points (34 GCPs). Products derived are Ortomosaic, point cloud, digital terrain model, digital surface model and contours lines.
The positional accuracy control process has been considered to check two sheets of the 1: 1000 grid (15 ”resolution), according to Image 6. The sample size for sheet 1 was 25 control points; and for sheet 2, of 24 control points.
Table 1 shows the set of coordinates obtained in the field, by GPS positioning (RTK method) and the respective orthophoto coordinates, for sheet 1 The spatial distribution of the controlled points and the representation of the magnitude of the positional error of sheet 1 is presented in Figure 1
The descriptive statistics of the horizontal positional error found in sheet 1 are presented in Table 2.
The mean positional error was 0.064 m. with a 95% confidence interval it is 0.064 ± 0.019 m.The positional accuracy at 95% confidence according to the NSSDA Standard (National Standard for Spatial Data Accuracy), based on the RMS, is 0.135 m. Table 3 shows the set of coordinates obtained in the field and in the orthophoto, for sheet 2.
The spatial distribution of the controlled points and the representation of the magnitude of the positional error of sheet 2 is presented in Figure 2.
The descriptive statistics of the horizontal positional error found in sheet 2 are presented in Table 4.
The mean positional error was 0.052 m. with a 95% confidence interval it is 0.052 ± 0.015 m. The positional accuracy at 95% confidence according to the NSSDA Standard (National Standard for Spatial Data Accuracy), based on the RMS, is 0.109 m.
From the analysis performed, based on the calculation of the 95% confidence interval of the mean positional error and the positional accuracy indicator according to the NSSDA Standard at the same level of confidence, for the two contracted letters of the 1: 1,000 grid, it can be concluded that they comply with the accuracies required for a 1: 1000 scale (compliance level of 30 cm.).
When the tolerance was verified, the drawing (Buildings, roads, blocks, trees, rivers, ..) was carried out according to the regulations of the military geographical institute, and the following steps to derive an urban cadastre.
Some people related with GIS area are skeptical of drone use as a tool for data measurement. Photogrammetry it is a tecnique designed years ago. Before they used a manned vehicle an metric camera. Well it is the same, now you have an autonomous vehicle and digital parametrized camera.If you have a stabled drone and you apply the tecniqle in a correct way , for sure you are going to have amazing results! Think about it, less time, less money, products in 3D. If this project has been done with a manned flight it represented 80% of the project cost, and with the drones It represented a cost of 30%. Just do it!
Modern photogrammetry Drone executed by Latitude Aerospace Solutions Drones and Pilot Designers: Ing. Pedro Meneses, Ing. Jorge Pantoja, Ing. José Barzallo. GCPs design, flight planning and processing photogrammetric data: Ing. Bethania Peña. Ing. Pedro Meneses Technical report:Ing. Bethania Peña GCPs Measurements and processing:Ing. Cesar Leiva Control of Positional Accuracy executed by the company Leiva Ingenieria Measurement of checkpoints and technical report:Ing. Cesar Leiva