Resolution is a key characteristic in imagery and quite often the most important choice to be made. So, what is resolution and which resolution do I need?
The spatial resolution determines the level of detail that can be seen in, or extracted from, imagery. A user of imagery may, for example, wish to be able to detect and then map cars, individual buildings or trees. On the other hand, another user may wish to detect forests and fields and to classify the land cover accordingly, but not need to detect the individual plants and trees. Selecting the appropriate spatial resolution is very important because, in general, the cost of imagery per square kilometre increases rapidly as more detail can be detected. To help you to see the impact of image resolution on the visual appearance of different features in imagery, we have developed an application using data from our partner Hexagon Geospatial, which you can access here.
A thematic overview of the relationship between resolution, imaging platform, coverage and price is shown below.
The NIIRS scale has been used in the interpretation of aerial imagery to help identify the appropriate type of imagery for different applications. More detail and examples of what can be extracted from imagery with different NIIRS levels can be found here.
As a rule of thumb, it is possible to work out the imagery resolution that you need, by considering the smallest (narrowest) dimension of the smallest object that you wish to detect in the imagery (detectable size). You then halve this value to get the (raster) resolution of the imagery, with some examples shown below (with map-scale) courtesy of Rajinder Nagi of ESRI.
If you know the map-scale that you need, then divide the map-scale by 2000 and this gives you the worst-case resolution that you need (1:50,000 would translate to 25m imagery or better).
Please note that this is simplified and there are many complexities that need to be taken into account for advanced uses of imagery, but the above serves as useful starting point for working out what imagery resolution means and how you can work out what you will need.
It is important to note that when you are ordering imagery, a larger imagery viewing angle (particularly of 20 degrees or more away from the vertical) will result in some degradation both of the positioning accuracy of the imagery and the image resolution (level of detail detectable in the imagery). This can be a "cost" of accepting a wider viewing angle in order to ensure more imaging opportunities of a location from a particular satellite or constellation.