Low Resolution Satellite Imagery

[yzg26]

Hello,

I am trying to produce dsm with low resolution satellite images (2.5 meters). I obtained a good solution when I worked on high resolution (0.5 meter pan band) images on the same satellite. However, when I want to work on multispectral bands on the same satellite (2.5 meters), DSM is not successful. I tried to change some related parameters in the config file and in the codes, but I could not get a successful solution. I think the problem is "disparity min max values" and "sift matching points". What would you suggest me for a solution? Thank you.

[l333wang]

I've tried to use multispectral imagery, while it cannot work. So I change the multispectral to one single band in greyscale, the code works well. I guess this code cannot work with multispectral images. Is there anywhere showed that s2p can work with multispectral? Am I missed something?

[gfacciol]

Hi, to obtain better results you can average the spectral bands before calling s2p. But feeding multiple bands does not improve the quality of the results and multiplies by a factor 3 the computation time, so we removed that functionality some years ago.

…

On Wed, Mar 29, 2023 at 8:16 PM Lanying Wang ***@***.***> wrote: I've tried to use multispectral imagery, while it cannot work. So I change the multispectral to one single band in greyscale, the code works well. I guess this code cannot work with multispectral images. Is there anywhere showed that s2p can work with multispectral? Am I missed something? — Reply to this email directly, view it on GitHub <#126 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABGFI2VLP3MQQUYIT4RLI3DW6R4BBANCNFSM52SHMMNQ> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>