Ocean object detection using SNAP-CLI

development
#1

Hi,

I am a newbie to Snap and working with it for Master Thesis.
I need to deal with ship detection or ocean object detection.
I found this tutorial: http://www.digital-geography.com/extracting-information-sentinel-1/

But I am not sure how to implement it programmatically (I am fine with Java or Python)
Or at least doing it using SNAP-CLI

Assuming that I have already download sentinel data but mass data (one specific date time but different region).

Thanks in advance :blush:

#2

Your best way to do this is to use SNAP’s Graph Processing Tool. I don’t see a need to write code.
Have a look at the Graph Builder and after that at the Batch Processing. I guess you will also find the help entry on Ocean Object Detection useful.

#3

Thanks Tonio, the point is when I check gpt of snap-cli. I can only find Object discrimination op. I am not sure which one is used for Ocean object detection.

#4

There is not a single operator for ocean object detection. Instead, it is a combination of multiple ones.
Open SNAP Desktop and look at Radar->SAR Applications->Ocean Applications->Ocean Object Detection (this is also what is meant in the tutorial). You will see there one tab per used operator. You can find each of these operators on the command line. My advice is still to use the GraphBuilder to create a graph, you can then run it from the command line.

1 Like
#5

Thank you. Btw, I am having problem with trying processing file using SNAP on Linux ubuntu. SNAP always crash when I process data with error log “Java insufficient memory”. I checked my machine and it has at least 3.0GB ram free at the time it runs SNAP.

Thanks in advance.