How to create an Average Image from Gray Scale (8 bits) JPEG Images

Question

I�m trying to create a mean image from 9 JPEG Images, but this mean image has lots of "noise", is distorced, and doesnt�resemble at all the original ones.
I need this mean image so I�ll correlate each of the original ones to find out the best image I�m going to choose ussing NCC (normalized cross correlation coeficient).
When I use setRGB(x,y,rgb), it returns diferent "images" according to the number of zeros, for ex.: rgb = 028028028 is different from 28028028 and from 282828. I can�t find a way to work it out.
My algorithm is the folowing:
        File f;
        double [] [] pixels = null;
        Color [] [] tempPixel = null;
        Integer [] [] byte0 = null;
        Integer [] [] byte1 = null;
        Integer [] [] byte2 = null;
        Integer [] [] byte3 = null;
     Integer [] [] tempPixels = null;
        mean = null;
        int w,h;
        fileNames = new String[numFrames];
        for (int i=1; i<numFrames; i++){
            fileName = path + "Frame#" + i + ".jpg";
            fileNames[i-1] = fileName;
        // just for getting width, height, for configuring pixels array and
        // buffered images - mean & best.
        fileName = path + "Frame#8.jpg";
        f= new File(fileName);
        try{
            temp = ImageIO.read(f);
        }catch(IOException io){ return false;}
        w = temp.getWidth();
        h = temp.getHeight();
        mean = new BufferedImage(w, h, BufferedImage.TYPE_BYTE_GRAY);
     pixels = new double [w][h];
        byte0 = new Integer [w][h];
        byte1 = new Integer [w][h];
        byte2 = new Integer [w][h];
        byte3 = new Integer [w][h];
        tempPixel = new Color[w][h];
          tempPixels = new Integer[w][h];
        f=null; temp = null;
        //reset pixels values.
        for(int x = 0; x < w; x++){
            for(int y = 0; y < h; y++){
                byte0[x][y] =0;
                byte1[x][y]=0;
                byte2[x][y]=0;
                byte3[x][y] = 0;
                pixels[x][y] = 0;
        // read each file (image), and according to W & H and collects the
        // pixels values.
        for(int j=0; j<numFrames-1; j++) {
            f = new File(fileNames[j]);
            try{
                temp = ImageIO.read(f);
            }catch(IOException io){ return false;}
            if(temp!=null)
                for(int x=0 ; x< temp.getWidth(); x++){
                for(int y=0; y<temp.getHeight(); y++){
                    tempPixel[x][y] = new Color(temp.getRGB(x,y));
              pixels[x][y] = pixels[x][y] + temp.getRGB(x,y) / 16581375;
                    //BLUE
                    byte0[x][y] = byte0[x][y] + (tempPixel[x][y].getBlue());//tempPixel[x][y].getBlue(); //& 0xff;
                    //GREEN
                    byte1[x][y] = byte1[x][y] + (tempPixel[x][y].getGreen());// >>8 & 0xff;
                    //RED
                    byte2[x][y] = byte2[x][y] + (tempPixel[x][y].getRed()); //>>16 & 0xff;
                    //ALPHA
                    byte3[x][y] = byte3[x][y] + (tempPixel[x][y].getAlpha()); //>>24 & 0xff;
            f=null;
        }//end of loop for j<numFrames
        String t0, t1, t2, t3;
        for(int x=0 ; x< temp.getWidth(); x++){
            for(int y=0; y< temp.getHeight(); y++){
                 byte0[x][y] = byte0[x][y] / numFrames;
                if((byte0[x][y])<10)
                         t0= "0" + byte0[x][y].toString();
                //else if((byte0[x][y])>=10 & ((byte0[x][y])<100))
                //    t0 = "0" + byte0[x][y].toString();
                else
                    t0 = byte0[x][y].toString();
                byte1[x][y] = byte1[x][y] / numFrames;
                if((byte1[x][y])<10)
                         t1=  "0" + byte1[x][y].toString();
                //else if((byte1[x][y])>=10 & ((byte1[x][y])<100))
                //    t1= "0" + byte1[x][y].toString();
                else
                    t1 = byte1[x][y].toString();
                byte2[x][y] = byte2[x][y] / numFrames;
                if((byte2[x][y])<10)
                         t2= "0" + byte2[x][y].toString();
                //else if((byte2[x][y])>=10 & ((byte2[x][y])<100))
                //    t2 = "0" + byte2[x][y].toString();
                else
                    t2 = byte2[x][y].toString();
                byte3[x][y] = byte3[x][y] / numFrames;
                if((byte1[x][y])<10)
                    t3="0" + byte3[x][y].toString();
                else
                    t3 = byte3[x][y].toString();
                    pixels[x][y] = (pixels[x][y]/numFrames) * 16581375;
                try{
                    String rgbs = t2+t1+t0;
                    Integer rgb = Integer.valueOf(rgbs);
                          //System.out.println("Valor do rgb: " +rgb);
                    mean.setRGB(x,y,(int)pixels[x][y]);
                }catch(NumberFormatException nfe){
                    System.out.println("ERROR: Integer to STRING: "  + nfe);
                    return false;
                }// end of catch

Answer 1

It�s ok, I�ve already found the solution.
Thanks.