Doubts of image quality after subsetting

Optical Toolbox
#1

Hello there,

I’ve downloaded Sentinel 2 data. I displayed the image in SNAP, then I resampled the data to 10m spatial resolution. After that, I subseted the image into a specific area. However, the quality of the image after being subseted is blurry and not sharp. Is it normal to happen like this? Or the problem has resolved?

#2

Hi fieranisa,

It depends, at which band are you looking, one of the original 10m or 60m bands?
Which resampling methods have you used?
Can you attach a screenshot of such a blurred image?

#3

The resampling method I’ve used was by reference band from source product and I’ve used Band 2 with 10m spatial resolution as the reference band. The method applied was by using the Nearest upsampling method.

This is the blurred image after subset.

image
image.png1161×862 223 KB

#4

Considering the lineament on the right is a road and the pixel spacing is 10m, I am not surprised that with this zoom level, the image seems blurry. Or did I miss something ?

#5

If you look at the same area with out the resampling I guess you will see the same effect. Also the scene looks a bit hazy. This intensifies the blurryness effect.
If you like, you could compare the images with QGIS for example.

This is a scene at similar zoom level. It looks also a bit blurry.

image
image.png971×725 121 KB

#6

Hi @fieranisa

Can you specify the acquistion details please: such as Orbit and Tile?

Thanks

Jan

MPC/CC Operations Manager

#7

The image obtained : S2B_MSIL1C_20180812T032529_N0206_R018_T47NQD_20180812T071019

Could you please check the quality of the image after resampled (before subset) and after subset…

#8

I see…all right… Thank you

#9

Hi @fieranisa
I’ll certainly check the input product.

Cheers

Jan

#10

Thanks in advance !

#11

Hello @fieranisa
I cannot see any multispectral misregistration in the image. Thus, the image is assessed as NOMINAL for quality.

In order to check, I have compared an S2B acquistion from March 2019 (it’s very difficult to find cloud-free images in your AOI - indeed the image has some thin cirrus and other cloud that gives the ‘rainbow effect’ in the bottom left corner :slight_smile: ). As you can see, the 2019 image exhibits the same behaviour as your 2018 product:

image
image.png1431×771 69.4 KB

(Just for interest, this is the loading pier of the Kapar Energy Ventures Power station near Selangor).

Cheers

Jan

1 Like
#12

All right… Understood… really appreciate your explanation and effort… Thank you so much ! =D

1 Like