Operations with images

Prev Tutorial: Mask operations on matrices

Next Tutorial: Adding (blending) two images using OpenCV

Input/Output

Images

Load an image from a file:

C++

        Mat img = imread(filename);

Java

            Mat img = Imgcodecs.imread(filename);

Python

    img = cv.imread(filename)

If you read a jpg file, a 3 channel image is created by default. If you need a grayscale image, use:

C++

        Mat img = imread(filename, IMREAD_GRAYSCALE);

Java

            Mat img = Imgcodecs.imread(filename, Imgcodecs.IMREAD_GRAYSCALE);

Python

    img = cv.imread(filename, cv.IMREAD_GRAYSCALE)

Note

Format of the file is determined by its content (first few bytes). To save an image to a file:

C++

        imwrite(filename, img);

Java

            Imgcodecs.imwrite(filename, img);

Python

    cv.imwrite(filename, img)

Note

Format of the file is determined by its extension.

Use cv::imdecode and cv::imencode to read and write an image from/to memory rather than a file.

Basic operations with images

Accessing pixel intensity values

In order to get pixel intensity value, you have to know the type of an image and the number of channels. Here is an example for a single channel grey scale image (type 8UC1) and pixel coordinates x and y:

C++

            Scalar intensity = img.at<uchar>(y, x);

Java

                byte[] imgData = new byte[(int) (img.total() * img.channels())];
                img.get(0, 0, imgData);
                byte intensity = imgData[y * img.cols() + x];

Python

    _intensity = img[y,x]

C++ version only: intensity.val[0] contains a value from 0 to 255. Note the ordering of x and y. Since in OpenCV images are represented by the same structure as matrices, we use the same convention for both cases - the 0-based row index (or y-coordinate) goes first and the 0-based column index (or x-coordinate) follows it. Alternatively, you can use the following notation (C++ only):

            Scalar intensity = img.at<uchar>(Point(x, y));

Now let us consider a 3 channel image with BGR color ordering (the default format returned by imread):

C++ code

            Vec3b intensity = img.at<Vec3b>(y, x);
            uchar blue = intensity.val[0];
            uchar green = intensity.val[1];
            uchar red = intensity.val[2];

Python Python

    _blue = img[y,x,0]
    _green = img[y,x,1]
    _red = img[y,x,2]

You can use the same method for floating-point images (for example, you can get such an image by running Sobel on a 3 channel image) (C++ only):

            Vec3f intensity = img.at<Vec3f>(y, x);
            float blue = intensity.val[0];
            float green = intensity.val[1];
            float red = intensity.val[2];

The same method can be used to change pixel intensities:

C++

            img.at<uchar>(y, x) = 128;

Java

                byte[] imgData = new byte[(int) (img.total() * img.channels())];
                imgData[y * img.cols() + x] = (byte) 128;
                img.put(0, 0, imgData);

Python

    img[y,x] = 128

There are functions in OpenCV, especially from calib3d module, such as cv::projectPoints, that take an array of 2D or 3D points in the form of Mat. Matrix should contain exactly one column, each row corresponds to a point, matrix type should be 32FC2 or 32FC3 correspondingly. Such a matrix can be easily constructed from std::vector (C++ only):

            vector<Point2f> points;
            //... fill the array
            Mat pointsMat = Mat(points);

One can access a point in this matrix using the same method Mat::at (C++ only):

            Point2f point = pointsMat.at<Point2f>(i, 0);

Memory management and reference counting

Mat is a structure that keeps matrix/image characteristics (rows and columns number, data type etc) and a pointer to data. So nothing prevents us from having several instances of Mat corresponding to the same data. A Mat keeps a reference count that tells if data has to be deallocated when a particular instance of Mat is destroyed. Here is an example of creating two matrices without copying data (C++ only):

        std::vector<Point3f> points;
        // .. fill the array
        Mat pointsMat = Mat(points).reshape(1);

As a result, we get a 32FC1 matrix with 3 columns instead of 32FC3 matrix with 1 column. pointsMat uses data from points and will not deallocate the memory when destroyed. In this particular instance, however, developer has to make sure that lifetime of points is longer than of pointsMat If we need to copy the data, this is done using, for example, cv::Mat::copyTo or cv::Mat::clone:

C++

        Mat img = imread("image.jpg");
        Mat img1 = img.clone();

Java

            Mat img = Imgcodecs.imread("image.jpg");
            Mat img1 = img.clone();

Python

    img = cv.imread('image.jpg')
    _img1 = np.copy(img)

To the contrary with C API where an output image had to be created by the developer, an empty output Mat can be supplied to each function. Each implementation calls Mat::create for a destination matrix. This method allocates data for a matrix if it is empty. If it is not empty and has the correct size and type, the method does nothing. If however, size or type are different from the input arguments, the data is deallocated (and lost) and a new data is allocated. For example:

C++

        Mat img = imread("image.jpg");
        Mat sobelx;
        Sobel(img, sobelx, CV_32F, 1, 0);

Java

            Mat img = Imgcodecs.imread("image.jpg");
            Mat sobelx = new Mat();
            Imgproc.Sobel(img, sobelx, CvType.CV_32F, 1, 0);

Python

    img = cv.imread('image.jpg')
    _sobelx = cv.Sobel(img, cv.CV_32F, 1, 0)

Primitive operations

There is a number of convenient operators defined on a matrix. For example, here is how we can make a black image from an existing greyscale image img

C++

            img = Scalar(0);

Java

                byte[] imgData = new byte[(int) (img.total() * img.channels())];
                Arrays.fill(imgData, (byte) 0);
                img.put(0, 0, imgData);

Python

    img[:] = 0

Selecting a region of interest:

C++

            Rect r(10, 10, 100, 100);
            Mat smallImg = img(r);

Java

                Rect r = new Rect(10, 10, 100, 100);
                Mat smallImg = img.submat(r);

Python

    _smallImg = img[10:110,10:110]

A conversion from Mat to C API data structures (C++ only):

        Mat img = imread("image.jpg");
        IplImage img1 = cvIplImage(img);
        CvMat m = cvMat(img);

Note that there is no data copying here.

Conversion from color to greyscale:

C++

        Mat img = imread("image.jpg"); // loading a 8UC3 image
        Mat grey;
        cvtColor(img, grey, COLOR_BGR2GRAY);

Java

            Mat img = Imgcodecs.imread("image.jpg"); // loading a 8UC3 image
            Mat grey = new Mat();
            Imgproc.cvtColor(img, grey, Imgproc.COLOR_BGR2GRAY);

Python

    img = cv.imread('image.jpg')
    _grey = cv.cvtColor(img, cv.COLOR_BGR2GRAY)

Change image type from 8UC1 to 32FC1:

C++

        src.convertTo(dst, CV_32F);

Java

            src.convertTo(dst, CvType.CV_32F);

Python

    _dst = src.astype(np.float32)

Visualizing images

It is very useful to see intermediate results of your algorithm during development process. OpenCV provides a convenient way of visualizing images. A 8U image can be shown using:

C++

        Mat img = imread("image.jpg");
        namedWindow("image", WINDOW_AUTOSIZE);
        imshow("image", img);
        waitKey();

Java

            Mat img = Imgcodecs.imread("image.jpg");
            HighGui.namedWindow("image", HighGui.WINDOW_AUTOSIZE);
            HighGui.imshow("image", img);
            HighGui.waitKey();

Python

    img = cv.imread('image.jpg')
    cv.namedWindow('image', cv.WINDOW_AUTOSIZE)
    cv.imshow('image', img)
    cv.waitKey()

A call to waitKey() starts a message passing cycle that waits for a key stroke in the "image" window. A 32F image needs to be converted to 8U type. For example:

C++

        Mat img = imread("image.jpg");
        Mat grey;
        cvtColor(img, grey, COLOR_BGR2GRAY);
        Mat sobelx;
        Sobel(grey, sobelx, CV_32F, 1, 0);
        double minVal, maxVal;
        minMaxLoc(sobelx, &minVal, &maxVal); //find minimum and maximum intensities
        Mat draw;
        sobelx.convertTo(draw, CV_8U, 255.0/(maxVal - minVal), -minVal * 255.0/(maxVal - minVal));
        namedWindow("image", WINDOW_AUTOSIZE);
        imshow("image", draw);
        waitKey();

Java

            Mat img = Imgcodecs.imread("image.jpg");
            Mat grey = new Mat();
            Imgproc.cvtColor(img, grey, Imgproc.COLOR_BGR2GRAY);
            Mat sobelx = new Mat();
            Imgproc.Sobel(grey, sobelx, CvType.CV_32F, 1, 0);
            MinMaxLocResult res = Core.minMaxLoc(sobelx); // find minimum and maximum intensities
            Mat draw = new Mat();
            double maxVal = res.maxVal, minVal = res.minVal;
            sobelx.convertTo(draw, CvType.CV_8U, 255.0 / (maxVal - minVal), -minVal * 255.0 / (maxVal - minVal));
            HighGui.namedWindow("image", HighGui.WINDOW_AUTOSIZE);
            HighGui.imshow("image", draw);
            HighGui.waitKey();

Python

    img = cv.imread('image.jpg')
    grey = cv.cvtColor(img, cv.COLOR_BGR2GRAY)
    sobelx = cv.Sobel(grey, cv.CV_32F, 1, 0)
    # find minimum and maximum intensities
    minVal = np.amin(sobelx)
    maxVal = np.amax(sobelx)
    draw = cv.convertScaleAbs(sobelx, alpha=255.0/(maxVal - minVal), beta=-minVal * 255.0/(maxVal - minVal))
    cv.namedWindow('image', cv.WINDOW_AUTOSIZE)
    cv.imshow('image', draw)
    cv.waitKey()

Note

Here cv::namedWindow is not necessary since it is immediately followed by cv::imshow. Nevertheless, it can be used to change the window properties or when using cv::createTrackbar