Hello everyone,
Can someone please let me know what does "ensemble average for a given window of pixels" mean when calculating a coherency or covariance matrix? I have taken this sentence from a paper "Polarimetric Target Decompositions and Light Gradient Boosting Machine for Crop Classification: A Comparative Evaluation".
According to my understanding I have to consider a window of size m x n and take the pixels of a radar image in that window and then perform ensemble averaging i.e. mean of those pixels. Please let me know whether my understanding is correct or not.
As of now I am performing a 2D convolution with a square window and varying their size to analyse the performance.
Thanks in advance.
Best regards
Pavan Kumar.
Simply to put, ensemble average is analogous to expected value or mean, in that it represents a sort of "average" for the stochastic process. In this case, the stochastic process is the scattering associated with distributed target surface. The ensemble average is different from the "time average", which is more like a typical average, in that it is the average value of a single outcome of a stochastic process. Hope this helps !
Sriram Saran Bhiravarasu Thanks for your reply. Do you have any implementation in python or MATLAB or C++ with you to acheive this objective? Because applying a box car filter with a square window is not useful with my implementation.
Dears Pavan and Sriram,
I have just arrived now to the same question. In some papers it appears the next operations for sentinel 1:
r = { |VV||SVH|}
where |x| means the modulus and {x} means ensemble average.
If I am working with GRD sentinel 1 image. I would take the absolute value of each band -VH and VV respectively- but from that point I do not know what to do as averaging VH and VV pixel by pixel sounds a bit rare to me.
What am I misunderstanding?
Kind regards,
Jaime
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing