How can i scale the servo motor from the info that i take from image?

Hello. I'm working on my graduation project. Simply, it's a MatLAB based human tracking with camera and servo motors. I drive servos(MG995 model) with Arduino. The algorithm that I use is basicly like that.

(Let's assume 30 FramePerSecond WebCam)

Take the one Frame -->Find the human face(if exist)-->Drive servos to human face at image

My problem occurs when using servos. No problem with the face finding code, but the servos do not work well. I use the CascadeObjector to find the face.

sp_mid=0.5;%first servo position
 integ=[0;0];
% PI constants for spin
kps=0.005; 
kis=0.005;
detection = vision.CascadeObjectDetector();
 bboxes = step(detection, frame);
    if ~isempty(bboxes)
       frame = insertObjectAnnotation(frame, 'rectangle', bboxes, 'Person','TextBoxOpacity', 0.4, 'FontSize', 10);
    if exist('bboxes')
        if ~isempty(bboxes)
          %(bboxes(1)+bboxes(3)/2) is the point of face on x-axis
    e=[ x_mid-(bboxes(1)+bboxes(3)/2);...  
                y_mid-(bboxes(2)+bboxes(4)/2)];
            % scaling
            e=e/10;  

            sp_val=sp_mid+kps*X_Point(1)+kis*integ(1);
               writePosition(spin,sp_val); 

How can I drive the servos to the face point that I find in the frame? My aim is to center the face at live video flow. Please, I need help.Thanks.

Answer 1

[Image Analyst]

I'd talk to the servo manufacturers to see if they have any MATLAB code to talk to their servos, especially through an Arduino.