How to form a feature vector?

Hello everyone I am doing a project "identification of person using retina". I have extracted my features (Bifurcation Points and end Points). Now I have to make a feature vector. This feature vector should have the distance and angles between four nearest features of every feature point. Please guide me about some methods through which I can do it. I am not asking for the code. I am only asking for some methods to do it. Below is the image of what I have done so far.

Answer 1

[KSSV]

As you have the positions (x,y) in your hand. You can use knnsearch to get the desired number of nearest neighbors. Once you have the nearest neighbors, you can find distance and angle. Read about knnsearch.

[Sidra Aleem]

@kSSV I have to find the distance and angle between four nearest feature points. Not the angle and distance between four nearest neighbors. I have given colors to feature point using Crossing number technique. This technique performs computation and depending on resulting value color is given.

Can you guide me a bit how can i find angle and distance.?

[KSSV]

If (x1,y1) and (x2,y2) are two points.

distance = sqrt((x2-x1)^2+(y2-y1)^2) ;
angle = atan2(y2-y1,x2-x1)