one.empty3.feature.jviolajones.Detector Maven / Gradle / Ivy
Show all versions of empty3-library-3d Show documentation
/*
* Copyright (c) 2022-2023. Manuel Daniel Dahmen
*
*
* Copyright 2012-2023 Manuel Daniel Dahmen
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package one.empty3.feature.jviolajones;
import org.jdom2.Document;
import org.jdom2.Element;
import org.jdom2.input.SAXBuilder;
import javax.imageio.ImageIO;
import java.awt.*;
import java.awt.image.BufferedImage;
import java.io.FileInputStream;
import java.io.InputStream;
import java.util.List;
import java.util.*;
public class Detector {
/*
* The list of classifiers that the test image should pass to be considered as an image.
*/
List stages;
Point size;
/*
* Detector constructor.
* Builds, from a XML file, the corresponding Haar cascade.
*
* @param filename The XML file (generated by OpenCV) describing the Haar cascade.
*/
public Detector(String filename) throws java.io.FileNotFoundException {
this(new FileInputStream(filename));
}
public Detector(InputStream input) {
Document document = null;
Element racine;
stages = new LinkedList();
SAXBuilder sxb = new SAXBuilder();
try {
//On cr�e un nouveau document JDOM avec en argument le fichier XML
//Le parsing est termin� ;)
document = sxb.build(input);
} catch (Exception e) {
e.printStackTrace();
}
//On initialise un nouvel element racine avec l'element racine du document.
racine = (Element) document.getRootElement().getChildren().get(0);
Scanner scanner = new Scanner(racine.getChild("size").getText());
size = new Point(scanner.nextInt(), scanner.nextInt());
Iterator it = racine.getChild("stages").getChildren("_").iterator();
while (it.hasNext()) {
Element stage = (Element) it.next();
float thres = Float.parseFloat(stage.getChild("stage_threshold").getText());
//Logger.getAnonymousLogger().log(Level.INFO, thres);
Iterator it2 = stage.getChild("trees").getChildren("_").iterator();
Stage st = new Stage(thres);
while (it2.hasNext()) {
Element tree = ((Element) it2.next());
Tree t = new Tree();
Iterator it4 = tree.getChildren("_").iterator();
while (it4.hasNext()) {
Element feature = (Element) it4.next();
float thres2 = Float.parseFloat(feature.getChild("threshold").getText());
int left_node = -1;
float left_val = 0;
boolean has_left_val = false;
int right_node = -1;
float right_val = 0;
boolean has_right_val = false;
Element e;
if ((e = feature.getChild("left_val")) != null) {
left_val = Float.parseFloat(e.getText());
has_left_val = true;
} else {
left_node = Integer.parseInt(feature.getChild("left_node").getText());
has_left_val = false;
}
if ((e = feature.getChild("right_val")) != null) {
right_val = Float.parseFloat(e.getText());
has_right_val = true;
} else {
right_node = Integer.parseInt(feature.getChild("right_node").getText());
has_right_val = false;
}
Feature f = new Feature(thres2, left_val, left_node, has_left_val, right_val, right_node, has_right_val, size);
Iterator it3 = feature.getChild("feature").getChild("rects").getChildren("_").iterator();
while (it3.hasNext()) {
String s = ((Element) it3.next()).getText().trim();
//Logger.getAnonymousLogger().log(Level.INFO, s);
Rect r = Rect.fromString(s);
f.add(r);
}
t.addFeature(f);
}
st.addTree(t);
//Logger.getAnonymousLogger().log(Level.INFO, "Number of nodes in tree "+t.features.size());
}
//Logger.getAnonymousLogger().log(Level.INFO, "Number of trees : "+ st.trees.size());
stages.add(st);
}
//Logger.getAnonymousLogger().log(Level.INFO, stages.size());
}
public org.jdom2.Element getChild(Element element, String childName) {
return null;
}
public List getFaces(String filename, float baseScale, float scale_inc, float increment, int min_neighbors, boolean doCannyPruning) throws java.io.FileNotFoundException, java.io.IOException {
return getFaces(new FileInputStream(filename), baseScale, scale_inc, increment, min_neighbors, doCannyPruning);
}
public List getFaces(InputStream input, float baseScale, float scale_inc, float increment, int min_neighbors, boolean doCannyPruning) throws java.io.FileNotFoundException, java.io.IOException {
return getFaces(ImageIO.read(input), baseScale, scale_inc, increment, min_neighbors, doCannyPruning);
}
/*
* Returns the list of detected objects in an image applying the Viola-Jones algorithm.
*
* The algorithm tests, from sliding windows on the image, of variable size, which regions should be considered as searched objects.
* Please see Wikipedia for a description of the algorithm.
*
* @param image bufferedimage input
* @param baseScale The initial ratio between the window size and the Haar classifier size (default 2).
* @param scale_inc The scale increment of the window size, at each step (default 1.25).
* @param increment The shift of the window at each sub-step, in terms of percentage of the window size.
* @return the list of rectangles containing searched objects, expressed in pixels.
*/
public List getFaces(BufferedImage image, float baseScale, float scale_inc, float increment, int min_neighbors, boolean doCannyPruning) {
List ret = new ArrayList();
int width = image.getWidth();
int height = image.getHeight();
float maxScale = (Math.min((width + 0.f) / size.x, (height + 0.0f) / size.y));
int[][] grayImage = new int[width][height];
int[][] img = new int[width][height];
int[][] squares = new int[width][height];
for (int i = 0; i < width; i++) {
int col = 0;
int col2 = 0;
for (int j = 0; j < height; j++) {
int c = image.getRGB(i, j);
int red = (c & 0x00ff0000) >> 16;
int green = (c & 0x0000ff00) >> 8;
int blue = c & 0x000000ff;
int value = (30 * red + 59 * green + 11 * blue) / 100;
img[i][j] = value;
grayImage[i][j] = (i > 0 ? grayImage[i - 1][j] : 0) + col + value;
squares[i][j] = (i > 0 ? squares[i - 1][j] : 0) + col2 + value * value;
col += value;
col2 += value * value;
}
}
int[][] canny = null;
if (doCannyPruning)
canny = getIntegralCanny(img);
for (float scale = baseScale; scale < maxScale; scale *= scale_inc) {
int step = (int) (scale * 24 * increment);
int size = (int) (scale * 24);
for (int i = 0; i < width - size; i += step) {
for (int j = 0; j < height - size; j += step) {
if (doCannyPruning) {
int edges_density = canny[i + size][j + size] + canny[i][j] - canny[i][j + size] - canny[i + size][j];
int d = edges_density / size / size;
if (d < 20 || d > 100)
continue;
}
boolean pass = true;
int k = 0;
for (Stage s : stages) {
if (!s.pass(grayImage, squares, i, j, scale)) {
pass = false;
//Logger.getAnonymousLogger().log(Level.INFO, "Failed at Stage "+k);
break;
}
k++;
}
if (pass) ret.add(new Rectangle(i, j, size, size));
}
}
}
return merge(ret, min_neighbors);
}
public int[][] getIntegralCanny(int[][] grayImage) {
int[][] canny = new int[grayImage.length][grayImage[0].length];
for (int i = 2; i < canny.length - 2; i++)
for (int j = 2; j < canny[0].length - 2; j++) {
int sum = 0;
sum += 2 * grayImage[i - 2][j - 2];
sum += 4 * grayImage[i - 2][j - 1];
sum += 5 * grayImage[i - 2][j + 0];
sum += 4 * grayImage[i - 2][j + 1];
sum += 2 * grayImage[i - 2][j + 2];
sum += 4 * grayImage[i - 1][j - 2];
sum += 9 * grayImage[i - 1][j - 1];
sum += 12 * grayImage[i - 1][j + 0];
sum += 9 * grayImage[i - 1][j + 1];
sum += 4 * grayImage[i - 1][j + 2];
sum += 5 * grayImage[i + 0][j - 2];
sum += 12 * grayImage[i + 0][j - 1];
sum += 15 * grayImage[i + 0][j + 0];
sum += 12 * grayImage[i + 0][j + 1];
sum += 5 * grayImage[i + 0][j + 2];
sum += 4 * grayImage[i + 1][j - 2];
sum += 9 * grayImage[i + 1][j - 1];
sum += 12 * grayImage[i + 1][j + 0];
sum += 9 * grayImage[i + 1][j + 1];
sum += 4 * grayImage[i + 1][j + 2];
sum += 2 * grayImage[i + 2][j - 2];
sum += 4 * grayImage[i + 2][j - 1];
sum += 5 * grayImage[i + 2][j + 0];
sum += 4 * grayImage[i + 2][j + 1];
sum += 2 * grayImage[i + 2][j + 2];
canny[i][j] = sum / 159;
//Logger.getAnonymousLogger().log(Level.INFO, canny[i][j]);
}
int[][] grad = new int[grayImage.length][grayImage[0].length];
for (int i = 1; i < canny.length - 1; i++)
for (int j = 1; j < canny[0].length - 1; j++) {
int grad_x = -canny[i - 1][j - 1] + canny[i + 1][j - 1] - 2 * canny[i - 1][j] + 2 * canny[i + 1][j] - canny[i - 1][j + 1] + canny[i + 1][j + 1];
int grad_y = canny[i - 1][j - 1] + 2 * canny[i][j - 1] + canny[i + 1][j - 1] - canny[i - 1][j + 1] - 2 * canny[i][j + 1] - canny[i + 1][j + 1];
grad[i][j] = Math.abs(grad_x) + Math.abs(grad_y);
//Logger.getAnonymousLogger().log(Level.INFO, grad[i][j]);
}
//JFrame f = new JFrame();
//f.setContentPane(new DessinChiffre(grad));
//f.setVisible(true);
for (int i = 0; i < canny.length; i++) {
int col = 0;
for (int j = 0; j < canny[0].length; j++) {
int value = grad[i][j];
canny[i][j] = (i > 0 ? canny[i - 1][j] : 0) + col + value;
col += value;
}
}
return canny;
}
public List merge(List rects, int min_neighbors) {
List retour = new LinkedList();
int[] ret = new int[rects.size()];
int nb_classes = 0;
for (int i = 0; i < rects.size(); i++) {
boolean found = false;
for (int j = 0; j < i; j++) {
if (equals(rects.get(j), rects.get(i))) {
found = true;
ret[i] = ret[j];
}
}
if (!found) {
ret[i] = nb_classes;
nb_classes++;
}
}
//Logger.getAnonymousLogger().log(Level.INFO, Arrays.toString(ret));
int[] neighbors = new int[nb_classes];
Rectangle[] rect = new Rectangle[nb_classes];
for (int i = 0; i < nb_classes; i++) {
neighbors[i] = 0;
rect[i] = new Rectangle(0, 0, 0, 0);
}
for (int i = 0; i < rects.size(); i++) {
neighbors[ret[i]]++;
rect[ret[i]].x += rects.get(i).x;
rect[ret[i]].y += rects.get(i).y;
rect[ret[i]].height += rects.get(i).height;
rect[ret[i]].width += rects.get(i).width;
}
for (int i = 0; i < nb_classes; i++) {
int n = neighbors[i];
if (n >= min_neighbors) {
Rectangle r = new Rectangle(0, 0, 0, 0);
r.x = (rect[i].x * 2 + n) / (2 * n);
r.y = (rect[i].y * 2 + n) / (2 * n);
r.width = (rect[i].width * 2 + n) / (2 * n);
r.height = (rect[i].height * 2 + n) / (2 * n);
retour.add(r);
}
}
return retour;
}
public boolean equals(Rectangle r1, Rectangle r2) {
int distance = (int) (r1.width * 0.2);
/*return r2.x <= r1.x + distance &&
r2.x >= r1.x - distance &&
r2.y <= r1.y + distance &&
r2.y >= r1.y - distance &&
r2.width <= (int)( r1.width * 1.2 ) &&
(int)( r2.width * 1.2 ) >= r1.width;*/
if (r2.x <= r1.x + distance &&
r2.x >= r1.x - distance &&
r2.y <= r1.y + distance &&
r2.y >= r1.y - distance &&
r2.width <= (int) (r1.width * 1.2) &&
(int) (r2.width * 1.2) >= r1.width) return true;
if (r1.x >= r2.x && r1.x + r1.width <= r2.x + r2.width && r1.y >= r2.y && r1.y + r1.height <= r2.y + r2.height)
return true;
return false;
}
}