Esempio n. 1
0
  /**
   * Start the mosaic generation, called from GUI.
   *
   * @param image Image to apply mosaic to.
   * @param density Magnitude of density of points in distribution.
   * @param type Type of distribution, Random or Uniform.
   * @param pointsOnly Whether to show only distribution points.
   */
  public void start(BufferedImage image, int density, String type, boolean pointsOnly) {

    // numPoints calculations are based on the idea of the fractional area
    // the fractional area is the theoretical area surrounding a uniform
    // distribution of points in the area of the image. The number represents
    // dividing the area of the image by the number of points distributed
    // in the area. By dividing area by the fractional area, you get the number
    // of points needed to achieve that density. fractional area ranges from
    // ~250 to 2500
    double fractionalArea = 0;
    if (density >= 1 && density <= 5) {
      fractionalArea = 2500 - (FRAC_AREA_STEP * (density - 1));
    } else {
      fractionalArea = DEFAULT_FRAC_AREA;
    }

    if (type.equalsIgnoreCase("RANDOM")) {
      // determine number of points as whole area / frac area

      int numPoints = (int) (image.getWidth() * image.getHeight() / fractionalArea);
      PlotTree tree = new RandomPlot(numPoints, image.getWidth(), image.getHeight());
      tesselation = new Tesselation(image, tree);

      // in the case of the poisson disc distribution, we want to determine density
      // as the distance between points. this is approximated by taking the square root
      // of the fractionalArea, because that approximates the distance between points
      // in a hypothetical distribution. The full transformation is:
      // numPoints = sqrt(w*h)/sqrt(w*h/fracArea)
      // numPoints^2 = w*h/(w*h/fracArea)
      // numPoints^2 = fracArea
      // numPoints = sqrt(fracArea)
    } else if (type.equalsIgnoreCase("UNIFORM")) {
      int numPoints = (int) Math.sqrt(fractionalArea);
      PlotTree tree = new PoissonPlot(numPoints, image.getWidth(), image.getHeight());
      tesselation = new Tesselation(image, tree);
    }

    tesselation.createDistribution();

    // Apply the distribution data either as a mosaic or points
    if (pointsOnly) {
      tesselation.drawPoints(Color.WHITE.getRGB());
    } else {
      tesselation.applyMosaic();
    }

    // free up resources and reset for another run
    // cleanUp();

  }
Esempio n. 2
0
 /**
  * To run after completing a run of processing. Nulls references to the data structures to save
  * space and prepare for another run.
  *
  * <p>Consider removing this and all cleanup code or improving
  */
 private void cleanUp() {
   tesselation.cleanUp();
   tesselation = null;
 }