The Influence of Quality Setting in SURE
With SURE you can choose to run a Project in Quality Ultra or High. The default Quality varies depending on the scenario.
Default (UAVs, Close Range)
The choice of a Scenario (necessary for the configuration of a SURE Project) implies a default Quality Setting. It is highly recommended to keep this value unchanged during production.
This article elaborates on how the Quality Setting reflects on the results and their production.
The different quality settings can be set in the GUI and CLI as described below:
The quality setting is located at the bottom of the processing panel. The quality level can be chosen from the drop-down menu:
The quality setting can be specified in the CLI using the following flag:
The Quality setting primarily refers to the resolution of the Input Images used to generate the output products. Quality Ultra generally means that the Input Images are used at their original (full) resolution. When choosing Quality level High, the Input Images are down-sampled once, effectively doubling the Ground Sampling Distance (GSD). In a similar manner, the resolution of the images used for texturing is also determined by the Quality setting. In short, the choice of Quality setting has an effect on the resolution of the products and total processing time.
Resolution of results
The level of geometric detail of the resulting reconstruction is dependent on the Quality setting chosen. The depth precision of the reconstructed points is maximized when quality Ultra is used. A higher depth precision leads to an increase in edge sharpness and generally more accuracy in the final results.
If no output GSD is explicitly set by the user, SURE estimates the nominal Ground Sampling Distance of the input images, at the selected quality level, in order to determine the pixel size of the output raster products (DSM, True Ortho).
Processing time and disk space requirement
The number of reconstructed points will naturally decrease as the Quality level is lowered
For generating raster products, we recommend to use a GSD parameter, that is equal or greater to the GSD of the Input Images used for Dense Matching.
The following table shows the general Quality setting recommendations based on the type of data collected and desired products.
Recommended Quality Setting
Large scale mapping with Nadir camera systems
City mapping with standard oblique camera systems (5-head configuration)
UAV projects. Optimal trade-off between quality and performance
City Mapping with custom camera systems or UAVs in which maximum output quality is required