The highgui module has been designed to provide an easy way to visualize the results and try the functionality of developed applications with OpenCV. As we saw in the previous chapter, this module supplies functions to perform the following operations:

Of course, the module contains more functions to enhance the user interaction with the software applications. Some of them will be explained in this chapter. In the following tbContrast code example, we can read an image file and two windows are created: the first one shows the original image and the other is the resulting image after increasing or decreasing the contrast to the original image applying a quite simple scaling operation. The following example shows how to create a trackbar in the window to easily change the contrast factor (scale) in the image. Let's see the code:

#include "opencv2/core/core.hpp"
#include "opencv2/highgui/highgui.hpp"
#include <iostream>

using namespace std;
using namespace cv;

int main(int argc, char* argv[]) {
    const char in_win[]="Orig. image";
    const char out_win[]="Image converted...(no saved)";
    int TBvalContrast=50; // Initial value of the TrackBar
    Mat out_img;

    if (argc != 2) {
        cout << "Usage: <cmd><input image_file>" << endl;
        return -1;
    }
    Mat in_img = imread(argv[1]); // Open and read the image
    if (in_img.empty()) {
        cout << "Error!!! Image cannot be loaded..." << endl;
        return -1;
    }
    namedWindow(in_win); // Creates window for orig. image
    moveWindow(in_win, 0, 0); // Move window to pos. (0, 0)
    imshow(in_win, in_img); // Shows original image
    namedWindow(out_win);
    createTrackbar("Contrast", out_win, &TBvalContrast, 100);
    cout << "Press Esc key to exit..." << endl;
    while (true) {
    in_img.convertTo(out_img, -1, TBvalContrast/50.0);
        imshow(out_win, out_img);
        if (waitKey(50) == 27) // If Esc key pressed breaks
            break;
    }
    return 0;
}

The following screenshot shows the original image (fruits.jpg) and the same image with increased contrast obtained with the tbContrast application.

Original image and the image with increased contrast

The code explanation is given as follows:

In the previous chapter, we saw code examples that can create an implicit infinite loop waiting for a pressed key with the function call waitKey (without arguments). The events on the application main window (for example, trackbars, mouse, and so on) are caught and handled inside of that loop. On the contrary, in this example, we create an infinite loop with a while statement applying the contrast change with the convertTo function with a scale factor from 0.0 (slider at 0) to 2.0 (slider at 100). The infinite loop breaks when the Esc key (ASCII code 27) is pressed. The implemented contrast method is quite simple because the new values for the pixels are calculated by multiplying the original value by a factor greater than 1.0 to increase contrast and a factor smaller than 1.0 to decrease contrast. In this method, when a pixel value exceeds 255 (in any channel), a rounding (saturate cast) has to be done.

Then, in the tbContrastCallB example, we show the same functionality, but using a trackbarcallback function that is called every time the slider is moved. Note that the events are handled when the waitKey function is called. The application ends if you press any key. The code is as follows:

//… (omitted for brevity)
#define IN_WIN "Orig. image"
#define OUT_WIN "Image converted...(no saved)"
Mat in_img, out_img;

// CallBack function for contrast TrackBar
void updateContrast(int TBvalContrast, void *userData=0) {

    in_img.convertTo(out_img, -1, TBvalContrast/50.0);
    imshow(OUT_WIN, out_img);
    return;
}

int main(int argc, char* argv[]) {

    int TBvalContrast=50; // Value of the TrackBar

    // (omitted for simplicity)
    in_img = imread(argv[1]); // Open and read the image
    // (omitted for simplicity)
    in_img.copyTo(out_img); // Copy orig. image to final img
    namedWindow(IN_WIN); // Creates window for orig. image
    moveWindow(IN_WIN, 0, 0); // Move window to pos. (0, 0)
    imshow(IN_WIN, in_img); // Shows original image
    namedWindow(OUT_WIN); // Creates window for converted image
    createTrackbar("Contrast", OUT_WIN, &TBvalContrast, 100,
                   updateContrast);
    imshow(OUT_WIN, out_img); // Shows converted image
    cout << "Press any key to exit..." << endl;
    waitKey();
    return 0;
}

In this example, a void pointer to the updatedContrast function is passed as argument to the createTrackbar function:

createTrackbar("Contrast", OUT_WIN, &TBvalContrast, 100,
updateContrast);

The callback function gets as its first argument the value of the slider in the trackbar and a void pointer to other user data. The new pixels values for the image will be calculated in this function.

Using a callback function cause a few changes in this new code because the accessible data inside this function has to be defined with global scope. Then, more complexity is avoided on the datatypes passed to the callback function as follows:

The two different implementations shown in the previous example produce the same results. However, it should be noted that after using a callback function, the resulting application (tbContrastCallB) is more efficient because the math operations for the image conversion only take place at the change of the trackbar slide (when the callback is executed). In the first version (tbContrast), the convertTo function is called inside the while loop even if the TBvalContrast variable doesn't change.