我正在尝试得到一个实时处理视频流来检测和像素化人脸。我有工作的代码跟踪面和画一个矩形周围的脸，我有一个单独的素描工作像素化的整个框架，但我不能想出如何约束像素化在面对矩形。任何关于如何将两者合并在一起的建议都将不胜感激。
我的面部跟踪代码是：

import processing.video.*;
import java.awt.Rectangle;
import gab.opencv.*;

OpenCV opencv;
Capture video;
Rectangle[] faces;

void setup() {

  size( 640, 480 );

  String[] cameras = Capture.list();
  video = new Capture(this, cameras[0]);
  opencv = new OpenCV(this, width, height);
  opencv.loadCascade(OpenCV.CASCADE_FRONTALFACE); 
  faces = opencv.detect();
  frameRate(24.0);
  video.start();
}

void draw() {

  opencv.loadImage(video);
  image(video, 0, 0);
  video.read();

  faces = opencv.detect(1.2, 3, 0, 40, 400);    //adjust last two numbers to give min and max for face size
  noFill();
  stroke(255, 20, 255);
  strokeWeight(2);
  for (int i = 0; i < faces.length; i++) {
    rect(faces[i].x, faces[i].y, faces[i].width, faces[i].height);
  }
}

我的像素代码是：

import processing.video.*;

int blockSize = 25;
int numPixelsWide, numPixelsHigh;

Capture video;

color movColors[];

void setup() {
  size(640, 480);
  noStroke();

  String[] cameras = Capture.list();
  video = new Capture(this, cameras[0]);
  video.start();

  numPixelsWide = width; 
  numPixelsHigh = height; 
  println(numPixelsWide);
  movColors = new color[numPixelsWide * numPixelsHigh];
}

void draw() {

  if (video.available() == true) {
    video.read();
    video.loadPixels();

    int count = 0;
    for (int j = 0; j < numPixelsHigh; j++) {
      for (int i = 0; i < numPixelsWide; i++) {
        movColors[count] = video.get(i*blockSize, j*blockSize);
        count++;
      }
    }
  }

  background(255);
  for (int j = 0; j < numPixelsHigh; j++) {
    for (int i = 0; i < numPixelsWide; i++) {
      fill(movColors[j*numPixelsWide + i]);
      rect(i*blockSize, j*blockSize, blockSize, blockSize);
    }
  }
}