Simple FMCW radar prediction filtering

I am attempting to filter data collectred by an FMCW radar. The output of the radar is in a CSV format that provides me with columns of data for x and y coordinates, target number, frame, and velocity. Essentially the target number is the number of detected objects per frame. For example frame 1 is repeated down the column 8 times since there are 8 targets detected. Each target obviosly has its own x and y values that I will be using.

My goal is to define objects based on if the targets are within a certain distance. From here I want to only plot objects if they are within a certain radius on the next frame. I am attempting to do this by starting on frame 2 and referencing frame 1.

I am getting a plot and I can also plot circles around my designated objects independently in each frame. I want to be able to plots objects that are defined based on the previous frame. Essentially if a group of dots is following a path in the next frame, then classify this as an object and plot.

Thanks

data=csvread('exp5.csv');
fps = 10;
xRange = 2;
yRange = 3;

delay = 1/fps; %[s] 
%Split csv into vector
x = data(:,4);
y = data(:,5);
frameNo = data(:,1);
tarTotal = data(:,2);
tarNo = data(:,3);
intensity = data(:,8);
intensityLOG = data(:,9);
velocity=data(:,8);
% previousframe=frameNo-1;
k = 1; 
%delay loop to play back in real time

%define values for the first frame
frameNo(1)=1;
x(1)=-0.0840;
y(1)=0.0215;
    

for j = 2:max(frameNo) %go through each frame starting at the second frame
    ok = k; %store k before loop
    xs = zeros(tarTotal(k),1);
    ys = zeros(tarTotal(k),1);

 
    for i = 1:tarTotal(k)
        scatter(x(k), y(k), 20, 'filled', 'k')
        hold on
        xlabel('X [m]')
        ylabel('Y [m]')
        xlim([-xRange,xRange])
        ylim([0,yRange])
        k = k+1;
         
        xs(i) = x(k);
        ys(i) = y(k);
    end
    
    for a = 1:size(xs) % define vector to hold points a 
        pointa = [xs(a) ys(a)];
        
        for b = 1:size(xs) % define vector to hold points b
            pointb = [xs(b) ys(b)];
            if pointa == pointb % ignore points that are on top of each other or repeated
                break;
            else
                dist(i) = abs(sqrt((pointb(1)-pointa(1))^2 + (pointb(2)-pointa(2))^2)); %calculate the distance between all the targets             
            end
        end
      
    end
    if dist(j)<=0.5     % if the distance between objects is small enough, define an object 
        objectcenter =  (pointa+pointb)/2; 
        predictdist(j)=dist(j)-dist(j-1); % attempting to relate the object centers between frames

    elseif predictdist(j)<2 % if the object center in the next frame is close to the first frame, plot a circle around the object
         viscircles(objectcenter,.1)
    end
    pause(delay)
    clf
end