This is required to generate the spectral indices in higher resolution.
Use the HPF resolution merge function of ERDAS IMAGINE software to combine high-resolution panchromatic data with lower resolution. or use image fusion in ILWIS to enhance spatial resolution.
The general method is called pansharpening. You can also do that in ENVI by applying one of the implemented methods e.g. gram-schmidt pansharpening. But be careful you only sharpen the bands that are covered by the pan band.
There are several techniques to do that based on automatic processes that you can always find under pansharpen category, but you can perform a more complex process based on PCA multivariate statistics, artificial intelligence, and kernel filtering under a program as Octave, R, or Scilab. However, you will have to consider that this procedure will always spoil the reflectance measurement of the layer.
Use the HPF resolution merge function of ERDAS IMAGINE software to combine high-resolution panchromatic data with lower resolution. or use image fusion in ILWIS to enhance spatial resolution.
Sadik Mahammad, You can try pansharpening in the Google Earth Engine. Ready-made code is here-
// Function to mask clouds using the quality band of Landsat 8.
var maskL8 = function(image) {
var qa = image.select('BQA');
/// Check that the cloud bit is off.
// See https://landsat.usgs.gov/collectionqualityband
var mask = qa.bitwiseAnd(1
I use the 'HPF' function of ERDAS IMAGINE to combine/merge high-resolution panchromatic data with low and medium resolution bands.
Another great option (and from my experiments it outperforms the afforementioned algorithms) is to use area-to-point regression Kriging (ATPRK) or area-to-point regression CoKriging (ATPRCoK). Just use linear regression where your covariate is the PAN band and downscale the regression residuals using ATPRK or ATPRCoK. Finally, simply add the regression prediction and the downscaled residuals to produce the downscaled image.
An example:
Assume you want to downscale the RED band (30m) and the PAN band (15m) is your covariate. I will assume that your CRS of your data is ESRI:54009 (World Mollweide), that is, you need a projected reference system (in meters). In R you can:
# load the necessary library
library(terra)
# set working directory (a folder where you have your data)
wd
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing