Advanced Kinect Depth Programming

This program uses the depth and RGB sensors together to create a 3D color image. You can use the arrow keys to move the image around. The vectorShading variable controls the “quality” of the image. The higher the value, the less pixels the program writes to the image and the less processing power the overall program takes.

import processing.opengl.*;

// import kinect library
import SimpleOpenNI.*;

// kinect object
SimpleOpenNI kinectObject;

// to store current kinect image
PImage rgbImage;
// default zoom of z axis
int ZoomAxisZ = 1000;

// change Z axis zoom
float zoomZAxis = 1;

// starting point for XY veiwing - in middle
float rotateXAxis = 1023/2;

// storage for all of the kinect depth data
PVector[] depthPoints;

// the current point that we will be drawing
PVector currentPoint;

// sets how many pixels to process and draw
int vectorShading = 10;
Setup method. Draws a window 1024 x 768 and enables kinect
depth and RGB sensor. Redraws both images to line up
with each other
void setup() {
 // creates a window 1024*768
 // and uses OPENGL 3D rendering
 size(1024, 768, OPENGL);

 // start a new kinect object
 kinectObject = new SimpleOpenNI(this);

 // enable the kinect depth sensor

 // enable the kinect color camera

 // align RGB and depth pictures together
Draw method. Updates kinect image, gets it, and translates the
image to view in draw window.
void draw() {
	 // draw the back ground black

	 // update the kinect image

	 // get the current image from the kinect
	 rgbImage = kinectObject.rgbImage();
	 // recrop image to half width size
	 // half height size and zoom back ZoomAxisZ
	 translate(width/2, height/2, -ZoomAxisZ);

	 // flip X axis around

	 // map rotateXAxis variable from -180 to 180
	 // then convert to radians and input to rotateY
	 rotateY(radians(map(rotateXAxis, 0, width, -180, 180)));

	 // after rotating X/Y rotate Z on center axis
	 // by pressing up and down
	 // get all depth vector points from kinect
	 depthPoints = kinectObject.depthMapRealWorld();

	 // for each depth point from the kinect get the
	 // current vector of that point and draw it
	 for (int i=0; i<depthPoints.length; i+=vectorShading) {

		 // get current depth point to draw
		 currentPoint = depthPoints[i];

		 // set the stroke color based on the color pixel

		 // color a point in the current vectors position
		 point(currentPoint.x, currentPoint.y, currentPoint.z);

Event driven method that moves image on X or Z axis to view
and prints keyCode for any key pressed for debugging.
void keyPressed(){
	 // if up is pressed then zoom in Z axis
	 if(keyCode == 38){
	 	zoomZAxis = zoomZAxis + 0.01;

	 // if down is pressed then zoon out Z axis
	 if(keyCode == 40){
	 	zoomZAxis = zoomZAxis - 0.01;

	 // if left is pressed then turn image left
	 if(keyCode == 37){
	 	rotateXAxis = rotateXAxis + 1;

	 // if right is pressed then turn image right
	 if(keyCode == 39){
	 	rotateXAxis = rotateXAxis - 1;

	 // print the current keyCode for the
	 // key you are pressing
If mouse is clicked in window then save image to where this
file is saved
void mousePressed(){
 	// save image
 	// print saved

Leave a Reply

Your email address will not be published.