Extracting mean values of numerous images in an easy way?

Hello all!

I am a long time follower of this forum and a huge admirer of you guys’/girls’ work in here. As a master student in remote sensing field and also as an employee of a disaster management authority of my country i learned a lot from you people and from this forum. And i am hoping you can help me out this time also.

As i mentioned before i am doing masters about flood analysis with GRD files.But the thing is i needed to split image to sub-images for a better analysis therefore i got a hundreds of sub-images(1600 to be exact)and i need their mean values as .csv or .txt so i can use them on excel later on. I know there is a way of getting the mean values of images by ‘analysis’ section one by one but for 1600 image it takes too much time .is there any easy way to do this easily?

P.s pls excuse my english if i did any msitakes its not my native tongue

What you try to achieve sounds like a job for the StatisticsOp. Unfortunately, this is not well documented in the current release. But a colleague of mine wrote a post which explains the usage.

Feel free to ask further questions if something is not clear in this description.

Thank you for the quick reply,

İ just looked over StaticsOP but couldnt really figured it. first off when i load the staticsOP.xml i get this screen

after that i just look at the .xml and find out need to apply some changes for my pc

so i made some changes in notepad++

but finally i still get the same notification

This operator is not working in the graph builder. You need to invoke it from the command line as explained in the post.

oh i see :grinning: again thank you for your quick reply. it seems statisticsOP is the tool i need and i feel like i am only one step away for what i need. Used statisticsOP sucessfully and get the results for 400 image. But the thing is mean value of 400 image had been calculated but i what i want is 400 mean value. mean value of each image. Made various changes at notepad++ but didnt changed the result

Unfortunately, it is not possible to run the StatisticsOperator with multiple products so it will output the mean of each product.
What you can do is to define temporal intervals and regions. If you know that there is an evenly spaced time interval between the products, you can define startdate, enddate and time interval (as in the example graph). Also, if you know that your products cover the same space, you can define these regions as vector data in a shapefile and pass this using the shapefile parameter. If you combine these approaches you will get a mean vector for each combination of time and interval.

However, I suppose in your case it makes most sense to apply batch processing. This will run the operator multiple times, every time on a single product. You will end up with one .csv-file per product, though. Here is a link to a tutorial on batch processing: http://www.brockmann-consult.de/beam-wiki/display/BEAM/Bulk+Processing+with+GPT