Picture

[Payx]

@VRonin Yes but in my .h i add QImage pixi thats not correct?

[VRonin]

I'm surprised your compiler doesn't yell about redefinition (I probably know why but that's not the point here). I'send you to http://www.cplusplus.com/doc/tutorial/namespaces/#scopes once again

[kshegunov]

@VRonin said in Picture:

I'm surprised your compiler doesn't yell about redefinition

Not redefinition, shadowing, hence no yelling. :)

[VRonin]

@kshegunov said in Picture:

shadowing

The case @kshegunov is explaining is in the 3rd code block of the link above:

// inner block scopes
#include <iostream>
using namespace std;

int main () {
  int x = 10;
  int y = 20;
  {
    int x;   // ok, inner scope.
    x = 50;  // sets value to inner x
    y = 50;  // sets value to (outer) y
    cout << "inner block:\n";
    cout << "x: " << x << '\n';
    cout << "y: " << y << '\n';
  }
  cout << "outer block:\n";
  cout << "x: " << x << '\n';
  cout << "y: " << y << '\n';
  return 0;
}

This means I should go back to that tutorial myself

[Payx]

@VRonin so if i want use my picture in an other function i have to declare it before my function as i did for

QColor dominantColor(const QImage& image, const QPoint& topLeft, const QSize& rectSize);

so i have to add a line code QImage pixi in my .cpp ?

why not in my .h, and why

QColor dominantColor

dont work in my .h too ?

[VRonin]

Ok, this has become too chatty, let's make it more structured before we both get banned.

• Put dominantColor and fillSection code from either me or @mrjj at the top of your .cpp file, just below the #includes no need to declare them in advance in this case
• in the private: section of your .h file add QImage m_baseImag;
• in on_push_clicked load the image from file and store it in m_baseImag then use QPixmap::fromImage(m_baseImag) to display the image in the QLabel
• in on_push2_clicked iterate over every 4x4 square of your picture calling the two functions:

for (int i=0;i<m_baseImag.width();i+=4){
           for (int j=0;j<m_baseImag.height();j+=4){
fillSection(m_baseImag,QPoint(i,j),QSize(4,4),dominantColour(m_baseImag,QPoint(i,j),QSize(4,4)));
}}

• use QPixmap::fromImage(m_baseImag) to display the new image in another QLabel

[Payx]

Hello,

I have 0 error but it didn't work.

here my code

void MainWindow::on_push_clicked()
{
	QString fileName = QFileDialog::getOpenFileName(this,
	tr("Open Image"), "/", tr("Image Files (*.png *.jpg *.bmp)"));
	QPixmap pix(fileName);
	ui->label->setPixmap(pix);
	const QSize s = pix.size();
	QImage pixi = QImage(pix.toImage());

	

}

void MainWindow::on_push2_clicked()
{
	QPoint point(0,0);
	QSize size(4,4);
	for(int i=0;i<pixi.width();i+=4){
		for(int j=0;j<pixi.height();j+=4){
			fillSection(pixi,point,size,dominantColor(pixi,point,size));
		}
	}

	QPixmap pixa=QPixmap::fromImage(pixi);

	ui->label_2->setPixmap(pixa);
}

In my .h i declared

private:
   Ui::MainWindow *ui;
   QImage pixi;

pixi is the name of my picture (= m_baseImag)

i display pixa (Qpix from pixi) in the same label as pixi (dont know if i can do that)

[mrjj]

Hi
Good work.
in here
void MainWindow::on_push_clicked()
{
......
QImage pixi = QImage(pix.toImage()); // This looks like a local variable? ( die by scope)
}

Did you mean to use the one from .H ?

so
pixi = QImage(pix.toImage());

[Payx]

i changed my code by using pixi = QImage(pix.toImage());

and in my .h i got QImage pixi.

I try to fix my rectSize to 15x15

And my picture at the end is the same, she changed a little (she got a blue pixel in the top left

[VRonin]

you never update point in your loop. it's always 0,0 it should be i,j

[Payx]

此回覆已被刪除！

[Payx]

Ok after long hours, i fixed it.

Thank you very much for your time in my project.

But i have a little question :

In the code it only returns Red, Blue or Green.

Do you have a function who return a lot of color, like if in my rect the most principal color is grey (for example) what could i do to add a lot of color ?

At the end i will have a picture more pixelated but that i can recognize

[mrjj]

Hi

• if in my rect the most principal color is grey

That is just values for R,G,B. still.

[VRonin]

Modify dominantColour with this one. It takes the most frequent colour that appears in the square and fills it with it

QColor dominantColour(const QImage& image,const QPoint& topLeft, const QSize& rectSize){
	const int maxRight = qMin(image.width(),topLeft.x() + rectSize.width());
	const int maxBottom = qMin(image.height(),topLeft.y() + rectSize.height());
	QHash<QColor,int> colourMap;
	for(int x=topLeft.x();x<maxRight;++x){
		for(int y=topLeft.y();y<maxBottom;++y){
			const QColor tempColor=image.pixelColor(x,y);
			auto colourIter=colourMap.find(tempColor);
			if(colourIter==colourMap.end())
				colourMap.insert(tempColor,1);
			else
				colourIter.value()++;
		}
	}
	if(colourMap.isEmpty())
		return Qt::black;
	return std::max_element(colourMap.constBegin(),colourMap.constEnd()).key();
}

[kshegunov]

I suggest you average the color in the rectangle, e.g. (thanks to @VRonin for the scaffolding):

QColor dominantColour(const QImage & image,const QPoint & topLeft, const QSize & rectSize)
{
    qint32 averageRed = 0, averageGreen = 0, averageBlue = 0;

    const qint32 maxX = qMin<qint32>(image.width(), topLeft.x() + rectSize.width());
    const qint32 maxY = qMin<qint32>(image.height(), topLeft.y() + rectSize.height());
    for (qint32 y = topLeft.y(); y < maxY; y++)  {
        for (qint32 x = topLeft.x(); x < maxX; x++)   {
            QRgb pixel = image.pixel(x, y);
            
            averageRed += qRed(pixel);
            averageGreen += qGreen(pixel);
            averageBlue += qBlue(pixel);
        }
    }

    qint32 n = rectSize.width() * rectSize.height();

    Q_ASSERT(n);
    if (n <= 0)
        return Qt::black;

    return QColor(averageRed / n, averageGreen / n, averageBlue / n);
}

[Payx]

Thank you very much !