SNAP multi-size mosaic processor too slow?

When I process 6 S2 zip files for the neigbouring datastrips of the same orbit with SNAP multisize mosaic processor, it will take extremely long time ,almost take 3 or more days, but similar mosaicking withENVI will take about 30 minutes or less than one hour (tested with 10m bands). In fact , I want to mosaic S2 data for more than 20 consecutive or neighbouring datastrips of the same orbit within 2~4 hours timespan, but it seems SNAP can’t sustain this processing load. Can anyone provide more efficient way or walkround about this requirement? Thanks.

Hi, I’m facing the same problem. Did you find any workaround?
Thanks

No, I haven’t found walkstrough or got tips yet. I hope SNAP development team can have an insight into the implementation of mosaicking processor ,or have a test on 20~30 consecutive images( each is the standard data frame of UTM 100km * 100km) to see how long it will take.

Hello, we performed a test on 25 images, you can see the duration below:

Regards
Adrian

How much memory are you giving SNAP? I suspect that it makes a huge difference in this case.

I’ve used the default SNAP memory configuration. The machine where I’ve run the Multi Size Mosaic has 32 GB memory and i7-10700 CPU with 16 logical cores.

Regards
Adrian

1 Like