welcome!

first thing: a better title helps with getting your qestions answered ..

i think you just need to know when the mouse-input starts and when it ends. in between you just record the coordinates in an array. once the user finished his drawing you can use that array for whatever you want ( send it to an object, keep it somewhere else, ..). so there is no need to use createGraphics which will just give you another "layer" for drawing.

maybe you need to describe your project in a little more detail for us .. so we can give better "support".

best,
F

something like:
Code:

boolean drawing = false;
Vector gestures;
Vector gesture;
void setup()
{
  size(800,600);
  gestures = new Vector();
  gesture = new Vector();
}
void draw()
{
  background(255);
  stroke(0);
  if(gesture != null){
  for(int i = 1; i < gesture.size(); i++){
    Point pp = (Point)gesture.get(i-1);
    Point p = (Point)gesture.get(i);
    line((float)pp.getX(),(float)pp.getY(),(float)p.getX(),(float)p.getY());
  }
  }
}

void setDrawing(boolean d)
{
  if(d){
    gesture = new Vector();
  }
  else{
    gestures.add(gesture);
  }
  drawing = d;
}

void mousePressed()
{
  setDrawing(!drawing);
}

void mouseDragged()
{
  gesture.add(new Point(mouseX,mouseY));
}

void mouseReleased()
{
  setDrawing(!drawing);
}

 

best
seltar