How to fit a 3D plane Pi on some binary image points using SVD and RANSAC and then finding the main axis by applying PCA on the obtained 3D points?

I am trying to implement a paper -

'Combining multiple depth-based descriptors for hand gesture

recognition' by Fabio Dominio, Mauro Donadeo, Pietro Zanuttigh

I am pasting a line from a paragraph under 'Hand Segmentation' of the paper (Page no. 104)-

"Once all the possible palm samples have been detected, we fit a

3D plane p on them by using SVD and RANSAC. Then Principal

Component Analysis (PCA) is applied to the 3D points in H in order

to extract the main axis that roughly corresponds to the direction

ix of the vector going from the wrist to the fingertips. "

According to the paper, I now have hand palm samples but I am not able to understand how to implement the above mentioned part. It is a big paragraph so I have pasted just the first few lines of it.

I have attached the paper PDF with this question. If anybody has any idea regarding the same whether in MATLAB or Python, please let me know.

Thanks in advance!

[KALYAN ACHARJYA]

Be specific please?

If the query is small I will try to figure out the answer.

[PRACHI SHARMA]

how to fit a 3D model π on some image points using Singular Value Decomposition (SVD) and Random Sample Consensus (RANSAC) using MATLAB or Python?