Webcam Video Capture

[bajoelkid12]

Hello everybody..

I'm new in Qt, and new in OpenCV..

I want to make a program to capture a video from USB webcam, then save it to the server..

For now, i have download a source code that use Qt and OpenCV to view images from webcam.. i download it here, http://qt-apps.org/content/show.php/...?content=89995

do you have any suggestion, how should i modifiy this source code, so that it can record the video, and save it ?

i really appreciate your help.. thanks before

P.S : sorry for my bad english

[stampede]

do you have any suggestion

Yes, read OpenCV documentation, for example:
CvVideoWriter
CvWriteFrame
Post again if you have specific problems with your code.

[bajoelkid12]

thank you for your respond stampede.. i'm now working on how to integrate OpenCV video writing and reading function in Qt framework.. . maybe you have some advice ?

[stampede]

You mean you are having some problems with using OpenCv in your code, or just thinking how to do that nicely ? If its the latter, then I'd create a wrapper class(es) with nice interface, which will hide all the OpenCV stuff from client classes. What I mean is something like:

class VideoReader : public QObject{
Q_OBJECT
public:
    VideoReader( const QString& file, QObject * parent = NULL );
    ... /// information
    int frameCount() const;
    QSize frameSize() const;
    ... // other methods, like video fps etc
    /// capture methods
    QImage getFrame( int frameNumber ) const; // or return QPixmap, or your own Image class
...
};

class VideoReader : public QObject{
Q_OBJECT
public:
    VideoReader( const QString& file, QObject * parent = NULL );
    ... /// information
    int frameCount() const;
    QSize frameSize() const;
    ... // other methods, like video fps etc
    /// capture methods
    QImage getFrame( int frameNumber ) const; // or return QPixmap, or your own Image class
...
};

To copy to clipboard, switch view to plain text mode 

It's always better to have class with good interface, than C-style OpenCv api calls all over the code. I think you get the idea.

[bajoelkid12]

actually yes.. i have some problems with using OpenCV in my code.. this is what i've done so far..

mycamerawindow.cpp

#include "MyCameraWindow.h"
 
MyCameraWindow::MyCameraWindow(CvCapture *cam, QWidget *parent) : QWidget(parent) {
    camera = cam;
    QVBoxLayout *layout = new QVBoxLayout;
    cvwidget = new QOpenCVWidget(this);
    layout->addWidget(cvwidget);
    setLayout(layout);
    resize(500, 400);
 
    startTimer(100);  // 0.1-second timer
 }
 
void MyCameraWindow::timerEvent(QTimerEvent*) {
    IplImage *image=cvQueryFrame(camera);
    cvwidget->putImage(image);
}

#include "MyCameraWindow.h"

MyCameraWindow::MyCameraWindow(CvCapture *cam, QWidget *parent) : QWidget(parent) {
    camera = cam;
    QVBoxLayout *layout = new QVBoxLayout;
    cvwidget = new QOpenCVWidget(this);
    layout->addWidget(cvwidget);
    setLayout(layout);
    resize(500, 400);

    startTimer(100);  // 0.1-second timer
 }

void MyCameraWindow::timerEvent(QTimerEvent*) {
    IplImage *image=cvQueryFrame(camera);
    cvwidget->putImage(image);
}

To copy to clipboard, switch view to plain text mode 

i get this code from http://qt-apps.org/content/show.php/...?content=89995

btw thank you for the algorithm, now i'll try to implement it to my code.. but if you don't mind, may i send my project file to you ?

[stampede]

actually yes.. i have some problems with using OpenCV in my code..

What kind of problems ? Image is not displayed ? Displayed but incorrect ? Other ?

but if you don't mind, may i send my project file to you ?

What for ? I think you can get all the help you need by asking specific questions on this forum.

[bajoelkid12]

actually.. i don't know how to start it.. i'm new in programming.. new in Qt and OpenCV..i've think about some algorithm that i'll use in my work, it's somewhat like :

1. Detect camera
2. If there is a camera, then initialize the camera
3. (the images from the camera are displayed)
4. then there will be some buttons for recording, pause, and stop functions
5. if the recording button is clicked, then save the recorded video into a spesific place and certain format.

that's all, and i just don't know how to implement those algorithm into codes..i don't know how to write them.. actually, this is my first experience..

Any advice Mr.Stampede ? i'm so sorry for the silly question from me..i really appreciate your help

---

Added after 6 minutes:

oh yes, for the basic need of integration (OpenCV2.2 and QtCreator 2010.05), i have read it here

http://www.barbato.us/2011/03/18/usi...s-2/#more-1625

now it's just "how to write the code" problems

[stampede]

If you are new both to Qt and OpenCV, then in my opinion best for you would be to write some simple separate applications using Qt and OpenCV, and start combining the two when you gain some experience with those libraries.

i'm new in programming..

In that case, I'd start with basic C/C++ before trying to code using Qt.

[bajoelkid12]

In that case, I'd start with basic C/C++ before trying to code using Qt

Do you mean TURBO C++ application ? or there is anyother ?

[squidge]

Yes, there is standard C++.

You can use Qt Creator to build C++ applications as well as Qt application. I would advise to learn C++ before Qt.

[bajoelkid12]

hello guys, i'm a bit confused on how to understand a following code

#include "QOpenCVWidget.h"
 
// Constructor
QOpenCVWidget::QOpenCVWidget(QWidget *parent) : QWidget(parent) {
    layout = new QVBoxLayout;
    imagelabel = new QLabel;
    QImage dummy(100,100,QImage::Format_RGB32);
    image = dummy;
    layout->addWidget(imagelabel);
    for (int x = 0; x < 100; x ++) {
        for (int y =0; y < 100; y++) {
            image.setPixel(x,y,qRgb(x, y, y));
        }
    }
    imagelabel->setPixmap(QPixmap::fromImage(image));
 
    setLayout(layout);
}
 
QOpenCVWidget::~QOpenCVWidget(void) {
 
}
 
void QOpenCVWidget::putImage(IplImage *cvimage) {
    int cvIndex, cvLineStart;
    // switch between bit depths
    switch (cvimage->depth) {
        case IPL_DEPTH_8U:
            switch (cvimage->nChannels) {
                case 3:
                    if ( (cvimage->width != image.width()) || (cvimage->height != image.height()) ) {
                        QImage temp(cvimage->width, cvimage->height, QImage::Format_RGB32);
                        image = temp;
                    }
                    cvIndex = 0; cvLineStart = 0;
                    for (int y = 0; y < cvimage->height; y++) {
                        unsigned char red,green,blue;
                        cvIndex = cvLineStart;
                        for (int x = 0; x < cvimage->width; x++) {
                            // DO it
                            red = cvimage->imageData[cvIndex+2];
                            green = cvimage->imageData[cvIndex+1];
                            blue = cvimage->imageData[cvIndex+0];
 
                            image.setPixel(x,y,qRgb(red, green, blue));
                            cvIndex += 3;
                        }
                        cvLineStart += cvimage->widthStep;
                    }
                    break;
                default:
                    printf("This number of channels is not supported\n");
                    break;
            }
            break;
        default:
            printf("This type of IplImage is not implemented in QOpenCVWidget\n");
            break;
    }
    imagelabel->setPixmap(QPixmap::fromImage(image));
}

#include "QOpenCVWidget.h"

// Constructor
QOpenCVWidget::QOpenCVWidget(QWidget *parent) : QWidget(parent) {
    layout = new QVBoxLayout;
    imagelabel = new QLabel;
    QImage dummy(100,100,QImage::Format_RGB32);
    image = dummy;
    layout->addWidget(imagelabel);
    for (int x = 0; x < 100; x ++) {
        for (int y =0; y < 100; y++) {
            image.setPixel(x,y,qRgb(x, y, y));
        }
    }
    imagelabel->setPixmap(QPixmap::fromImage(image));

    setLayout(layout);
}

QOpenCVWidget::~QOpenCVWidget(void) {

}

void QOpenCVWidget::putImage(IplImage *cvimage) {
    int cvIndex, cvLineStart;
    // switch between bit depths
    switch (cvimage->depth) {
        case IPL_DEPTH_8U:
            switch (cvimage->nChannels) {
                case 3:
                    if ( (cvimage->width != image.width()) || (cvimage->height != image.height()) ) {
                        QImage temp(cvimage->width, cvimage->height, QImage::Format_RGB32);
                        image = temp;
                    }
                    cvIndex = 0; cvLineStart = 0;
                    for (int y = 0; y < cvimage->height; y++) {
                        unsigned char red,green,blue;
                        cvIndex = cvLineStart;
                        for (int x = 0; x < cvimage->width; x++) {
                            // DO it
                            red = cvimage->imageData[cvIndex+2];
                            green = cvimage->imageData[cvIndex+1];
                            blue = cvimage->imageData[cvIndex+0];

                            image.setPixel(x,y,qRgb(red, green, blue));
                            cvIndex += 3;
                        }
                        cvLineStart += cvimage->widthStep;
                    }
                    break;
                default:
                    printf("This number of channels is not supported\n");
                    break;
            }
            break;
        default:
            printf("This type of IplImage is not implemented in QOpenCVWidget\n");
            break;
    }
    imagelabel->setPixmap(QPixmap::fromImage(image));
}

To copy to clipboard, switch view to plain text mode 

Can you help me ?

[high_flyer]

Its quite a bit of code, with which part of it do you have trouble exactly?

[bajoelkid12]

thank you for the response.. .

first of all, what is the function of this code in your opinion ?

in my opinion, this code is to draw an image wrapped in Qimage, with RGB32 format.. but i don't know actually the drawing process based on that code.. .

[high_flyer]

in my opinion, this code is to draw an image wrapped in Qimage, with RGB32 format..

yes, but not only, it seems to convert BGR to RGB as well.

for (int x = 0; x < cvimage->width; x++) {
                            // DO it
                            red = cvimage->imageData[cvIndex+2];
                            green = cvimage->imageData[cvIndex+1];
                            blue = cvimage->imageData[cvIndex+0];
 
                            image.setPixel(x,y,qRgb(red, green, blue));
                            cvIndex += 3;
                        }

for (int x = 0; x < cvimage->width; x++) {
                            // DO it
                            red = cvimage->imageData[cvIndex+2];
                            green = cvimage->imageData[cvIndex+1];
                            blue = cvimage->imageData[cvIndex+0];

                            image.setPixel(x,y,qRgb(red, green, blue));
                            cvIndex += 3;
                        }

To copy to clipboard, switch view to plain text mode 

[bajoelkid12]

Hello guys, now i want to ask about video compression.. .is there any library that can support video compression method ? such as mpeg, or avi ? i'm using OpenCV2.2 right now. does it support video compression too ? thanks

[high_flyer]

Why do you ask it here?
This is a Qt forum, not OpenCV froum.
Try reading the OpenCV documentation first.
http://opencv.willowgarage.com/wiki/VideoCodecs

Ans ask OpenCV related question on their forum. you are more likely to get better answers there for OpenCV related questions.

[bajoelkid12]

hello guys.. now i want to ask, how to build a client - server application to transfer images that captured by webcam to another part of the network ? any suggestion ? thanks.. .

[squidge]

Maybe QNetworkAccessManager, or just plain old QTcpSocket, depending on what your server requires.

[bajoelkid12]

my server require to read the video stream that was sent by client.. .what do you suggest, QTcpSocket or QUdpSocket ? thank you

[squidge]

Which does your server support?