i need to know how the minimum norm can be used to estimate the degaradation among the two images.
Hello,
You should precise what you mean by "another form". Is one image derived from the other (by any image processing), or is it the same scene acquired with different modalities and / or different view points...?
Best regards, Pascal
another form means its actually called initial estimate,where i have added the high frequency coefficients of the 1st image on to the 2nd image set. by adding the coefficients the second image changes to a noisy sort of image.
Sorry, it's not precise enough. For instance: What is the original content of image 2? any relationship with image 1? I understood the following but I2 is still unknown:
Inputs: I1, I2
Output: I2' = I1' + I2
Goal: compare I2 and I2'
As Pascal pointed out, it is not very clear what your problem is. But for image comparisons in general, there are several methods and metrics. To name a few: Peak-Signal-to-Noise-Ratio (PSNR), EdgePSNR, Structural SIMilarity (SSIM), Feature SIMilarity (FSIM) and ..... The MATLAB codes for these methods are available with a Google search.
I hope this helps.
@Pascal Bertolino the images are actually satellite images,where in satellite there are two kinds of images MS(multispectral) and PAN (panchromatic). In my experiments what i have done is i took contourlet transform over these images and as said before the high frequency components of PAN are substituted on to MS image. such that a new image is formed by this substitution. i just need to find degradation matrix using minimum norm technique considering it as a underdetermined system of equation.
the equation is y= az+n, where y is input MS image,A is the degradation matrix,z is the final formed image(substitution of coefficients) and finally n is the noise in the images.
- Badges
- Science topic