Howdy Thayalan,

For digital imagery this is a piece of cake. Many consumer camera's do that on the fly and as Giulio has suggested, many software packages contain standard routines to resample a digital image to another pixel resolution.

When you are dealing with satellite imagery, in principle the technique to change a satellite image resolution is the same and the software as well. But it's a one way track, unless your satellite camera has zooming capacity, which most if not all (except some military ones), don't have. Commercial camera's nowadays always have a zoom objective. Apparently the military want to keep this type of camera's on the earth's surface. They have always been a bit selfish no?

I mean you can change the resolution from high to low, but it is usless to do the reverse, because you will lose information. I guess that' this is not what you would like to do.

So what is your objective? Let me see the back of your tongue on that.

resampling imagery form different sensors to the same resolution maybe?

Let me know.

Thanks,

Frank

you mean pan-sharpening? use OSSIM

Yes pan sharpening is a standard technique or even if there is any data with slightly different pixel sizes it can be resampled to a specific pixel size.

Pan-sharpening can be done using different algorithms, the best choice depends on the data you have and on the purpose. There are methods which preserve as much as possible the spectral information (e.g. wavelet based methods) or methods which improve as much as possible the appearance of the iamge (usually based on the IHS space).

Concerning free tools for performing pansharpening, consider Orfeo Tool Box (http://www.orfeo-toolbox.org/otb/) which also embedded into QGis.

On the other hand, practically every commercial package like ENVI or ERDAS Imagine include plenty of algorithms to perform that.

Howdy Thayalan,

For digital imagery this is a piece of cake. Many consumer camera's do that on the fly and as Giulio has suggested, many software packages contain standard routines to resample a digital image to another pixel resolution.

When you are dealing with satellite imagery, in principle the technique to change a satellite image resolution is the same and the software as well. But it's a one way track, unless your satellite camera has zooming capacity, which most if not all (except some military ones), don't have. Commercial camera's nowadays always have a zoom objective. Apparently the military want to keep this type of camera's on the earth's surface. They have always been a bit selfish no?

I mean you can change the resolution from high to low, but it is usless to do the reverse, because you will lose information. I guess that' this is not what you would like to do.

So what is your objective? Let me see the back of your tongue on that.

resampling imagery form different sensors to the same resolution maybe?

Let me know.

Thanks,

Frank

You must take care about the problems of changing the pixels sizes, because the image may be altered too much and maybe can not achieve your expectations.

Thaylan, to complement what was told above, look for papers on the IHS transformation.

PS as you are planning to improve your resolution by nearly an order of magnitude, I would appreciate if you share your results at some point later

Will appreciate if we can see the results.

You can do it with some software like ENVI. But, be careful that during pan sharping you will miss some information because you manipulate the data.

If you don´t like to loss spectral properties, you can try the downscaling with a fussion method, may be Landsat and quickbird, I know many works with landsat an MODIS (Roy, 2008; Busseto, 2008 for instance) or Landsat-Meris(look for Julia Amorós papers).

Do you want to do it a long all the spectral range, o just in an restricted area?

you can try the down scaling with Resolution merge technique ef you have two different resolution data one is higher and other is lower.........

i have a hyperspectral image with 3m spatial resolution and have multispectral image with 1.6 m spatial resolution. I want to resample the hyperspectral image resolution to 6.4 m after i should give the projection for this image and do geometric correction taking the multispectral image as a reference. The goal is image fusion based deep learning for image classification. how i can resample the hyperspectral image in ENVI or ERDAS or ArcGis also i should minimize the loss of spectral information. Also, is this preprocessing of my data right for the process of image fusion ?

Due to the importance of high-resolution multi-spectral images in many remote sensing applications, pan-sharpening techniques have been proposed to increase the spatial resolution of a low-resolution multi-spectral (Hyper-spectral) image using a high-resolution panchromatic(multi-spectral) images. To get more information about these techniques you can see some detailed surveys such as:

“Hyperspectral pan-sharpening: A review”

“A Critical Comparison Among Pan-sharpening Algorithms “

“A review of remote sensing image fusion methods"

My MS thesis was about Multi-spectral pan-sharpening; in which we proposed two methods entitled “A Self-Learning Approach for Pan-sharpening of Multi-spectral Images” and “A Compressed-Sensing-Based Approach for Remote Sensing Image Fusion”.

Moreover, you can find some Multi-spectral pan-sharpening Matlab code in below link: http://openremotesensing.net/knowledgebase/a-critical-comparison-among-pansharpening-algorithms/

further, you can find some Hyper-spectral pan-sharpening Matlab code in below link:

http://openremotesensing.net/knowledgebase/hyperspectral-and-multispectral-data-fusion/

Thank you