Normalize hyperspectral image using a target/reference

I have a hyperspectral image with a target reference within the same image as the data I want to analyze. From what I understand I need to normalize the image using this reference before I start working with the data.

I think I found an example of normalizing hyperspectral data with on this page about denoising hyperspectral images: https://www.mathworks.com/help/images/ref/denoisengmeet.html

In the example on how to use the denoise function I think this is the line used to normalize the image: hcube = hypercube(rescale(hcube.DataCube),hcube.Wavelength);

I found the coordinates of the part of the target I want by using the hyperspectralViewer() function, but haven't found a way to use them as input for the above method of normalizing the data. Is there a way to do this, or do I have to use a different approach when I want to normalize using a reference?

Answer 1

[Prachi Kulkarni]

Hi,

If you want to normalize all the data in the hyperspectral image, such that-

• the minimum of the normalized hyperspectral image is same as the minimum of the reference region
• the maximum of the normalized hyperspectral image is same as the maximum of the reference region

you can use the sample code below

hcube = hypercube('paviaU.hdr');
data = hcube.DataCube;

x_range = 1:100; % example range
y_range = 1:50; % example range
z_range = 1:25; % example range
reference = data(x_range,y_range,z_range);
lower_bound = min(reference(:));
upper_bound = max(reference(:));

rescaled_data = rescale(data,lower_bound,upper_bound);
hcube = hypercube(rescaled_data,hcube.Wavelength);