The Use Case:

Request a thresholded NDVI from the server, say, for further analysis and masking in a GIS.

The service:

Top performance, people often say, is only possible with really simple services, such as WMTS. We contend, though, that service quality rules, and show this by the following comparison.

Left a false color image is stored a a file, and it is retrieved the fastest possible way, directly from the Apache Web server. Right the same file got ingested into rasdaman, and via a WCPS query we load it, compute the NDVI, and apply the threshold which you can set interactively.

Wait...file loading should be faster, but it isn't, although there is processing involved additionally! How come? Well, processing in rasdaman is fast enough to not realize it. What you do see, however, is data transport. Let's inspect in detail: left we transport 24 bit per pixel, right it is 1 bit - and the image obviously compresses nicely. Bottom line, we transport Megabytes versus Kilobytes. This dynamically generated WCPS query delivers the thresholded NDVI.

Lesson learnt: an intelligent server with rich functionality (i) can be faster and (ii) offers better service quality, it delivers results instead of just data.


Loaded in 0 seconds
Loaded in 0 seconds