OpenCV  3.2.0
Open Source Computer Vision
Image Segmentation with Distance Transform and Watershed Algorithm

.2.0+dfsg_doc_tutorials_imgproc_imgtrans_distance_transformation_distance_transform

Goal

In this tutorial you will learn how to:

  • Use the OpenCV function cv::filter2D in order to perform some laplacian filtering for image sharpening
  • Use the OpenCV function cv::distanceTransform in order to obtain the derived representation of a binary image, where the value of each pixel is replaced by its distance to the nearest background pixel
  • Use the OpenCV function cv::watershed in order to isolate objects in the image from the background

Theory

Code

This tutorial code's is shown lines below. You can also download it from here.

#include <opencv2/opencv.hpp>
#include <iostream>
using namespace std;
using namespace cv;
int main(int, char** argv)
{
// Load the image
Mat src = imread(argv[1]);
// Check if everything was fine
if (!src.data)
return -1;
// Show source image
imshow("Source Image", src);
// Change the background from white to black, since that will help later to extract
// better results during the use of Distance Transform
for( int x = 0; x < src.rows; x++ ) {
for( int y = 0; y < src.cols; y++ ) {
if ( src.at<Vec3b>(x, y) == Vec3b(255,255,255) ) {
src.at<Vec3b>(x, y)[0] = 0;
src.at<Vec3b>(x, y)[1] = 0;
src.at<Vec3b>(x, y)[2] = 0;
}
}
}
// Show output image
imshow("Black Background Image", src);
// Create a kernel that we will use for accuting/sharpening our image
Mat kernel = (Mat_<float>(3,3) <<
1, 1, 1,
1, -8, 1,
1, 1, 1); // an approximation of second derivative, a quite strong kernel
// do the laplacian filtering as it is
// well, we need to convert everything in something more deeper then CV_8U
// because the kernel has some negative values,
// and we can expect in general to have a Laplacian image with negative values
// BUT a 8bits unsigned int (the one we are working with) can contain values from 0 to 255
// so the possible negative number will be truncated
Mat imgLaplacian;
Mat sharp = src; // copy source image to another temporary one
filter2D(sharp, imgLaplacian, CV_32F, kernel);
src.convertTo(sharp, CV_32F);
Mat imgResult = sharp - imgLaplacian;
// convert back to 8bits gray scale
imgResult.convertTo(imgResult, CV_8UC3);
imgLaplacian.convertTo(imgLaplacian, CV_8UC3);
// imshow( "Laplace Filtered Image", imgLaplacian );
imshow( "New Sharped Image", imgResult );
src = imgResult; // copy back
// Create binary image from source image
Mat bw;
cvtColor(src, bw, CV_BGR2GRAY);
imshow("Binary Image", bw);
// Perform the distance transform algorithm
Mat dist;
// Normalize the distance image for range = {0.0, 1.0}
// so we can visualize and threshold it
normalize(dist, dist, 0, 1., NORM_MINMAX);
imshow("Distance Transform Image", dist);
// Threshold to obtain the peaks
// This will be the markers for the foreground objects
threshold(dist, dist, .4, 1., CV_THRESH_BINARY);
// Dilate a bit the dist image
Mat kernel1 = Mat::ones(3, 3, CV_8UC1);
dilate(dist, dist, kernel1);
imshow("Peaks", dist);
// Create the CV_8U version of the distance image
// It is needed for findContours()
Mat dist_8u;
dist.convertTo(dist_8u, CV_8U);
// Find total markers
vector<vector<Point> > contours;
// Create the marker image for the watershed algorithm
Mat markers = Mat::zeros(dist.size(), CV_32SC1);
// Draw the foreground markers
for (size_t i = 0; i < contours.size(); i++)
drawContours(markers, contours, static_cast<int>(i), Scalar::all(static_cast<int>(i)+1), -1);
// Draw the background marker
circle(markers, Point(5,5), 3, CV_RGB(255,255,255), -1);
imshow("Markers", markers*10000);
// Perform the watershed algorithm
watershed(src, markers);
Mat mark = Mat::zeros(markers.size(), CV_8UC1);
markers.convertTo(mark, CV_8UC1);
bitwise_not(mark, mark);
// imshow("Markers_v2", mark); // uncomment this if you want to see how the mark
// image looks like at that point
// Generate random colors
vector<Vec3b> colors;
for (size_t i = 0; i < contours.size(); i++)
{
int b = theRNG().uniform(0, 255);
int g = theRNG().uniform(0, 255);
int r = theRNG().uniform(0, 255);
colors.push_back(Vec3b((uchar)b, (uchar)g, (uchar)r));
}
// Create the result image
Mat dst = Mat::zeros(markers.size(), CV_8UC3);
// Fill labeled objects with random colors
for (int i = 0; i < markers.rows; i++)
{
for (int j = 0; j < markers.cols; j++)
{
int index = markers.at<int>(i,j);
if (index > 0 && index <= static_cast<int>(contours.size()))
dst.at<Vec3b>(i,j) = colors[index-1];
else
dst.at<Vec3b>(i,j) = Vec3b(0,0,0);
}
}
// Visualize the final image
imshow("Final Result", dst);
waitKey(0);
return 0;
}

Explanation / Result

  1. Load the source image and check if it is loaded without any problem, then show it:
    // Load the image
    Mat src = imread(argv[1]);
    // Check if everything was fine
    if (!src.data)
    return -1;
    // Show source image
    imshow("Source Image", src);
  2. Then if we have an image with white background, it is good to tranform it black. This will help us to desciminate the foreground objects easier when we will apply the Distance Transform:
    // Change the background from white to black, since that will help later to extract
    // better results during the use of Distance Transform
    for( int x = 0; x < src.rows; x++ ) {
    for( int y = 0; y < src.cols; y++ ) {
    if ( src.at<Vec3b>(x, y) == Vec3b(255,255,255) ) {
    src.at<Vec3b>(x, y)[0] = 0;
    src.at<Vec3b>(x, y)[1] = 0;
    src.at<Vec3b>(x, y)[2] = 0;
    }
    }
    }
    // Show output image
    imshow("Black Background Image", src);
  3. Afterwards we will sharp our image in order to acute the edges of the foreground objects. We will apply a laplacian filter with a quite strong filter (an approximation of second derivative):

    // Create a kernel that we will use for accuting/sharpening our image
    Mat kernel = (Mat_<float>(3,3) <<
    1, 1, 1,
    1, -8, 1,
    1, 1, 1); // an approximation of second derivative, a quite strong kernel
    // do the laplacian filtering as it is
    // well, we need to convert everything in something more deeper then CV_8U
    // because the kernel has some negative values,
    // and we can expect in general to have a Laplacian image with negative values
    // BUT a 8bits unsigned int (the one we are working with) can contain values from 0 to 255
    // so the possible negative number will be truncated
    Mat imgLaplacian;
    Mat sharp = src; // copy source image to another temporary one
    filter2D(sharp, imgLaplacian, CV_32F, kernel);
    src.convertTo(sharp, CV_32F);
    Mat imgResult = sharp - imgLaplacian;
    // convert back to 8bits gray scale
    imgResult.convertTo(imgResult, CV_8UC3);
    imgLaplacian.convertTo(imgLaplacian, CV_8UC3);
    // imshow( "Laplace Filtered Image", imgLaplacian );
    imshow( "New Sharped Image", imgResult );

  4. Now we tranfrom our new sharped source image to a grayscale and a binary one, respectively:
    // Create binary image from source image
    Mat bw;
    cvtColor(src, bw, CV_BGR2GRAY);
    imshow("Binary Image", bw);
  5. We are ready now to apply the Distance Tranform on the binary image. Moreover, we normalize the output image in order to be able visualize and threshold the result:
    // Perform the distance transform algorithm
    Mat dist;
    // Normalize the distance image for range = {0.0, 1.0}
    // so we can visualize and threshold it
    normalize(dist, dist, 0, 1., NORM_MINMAX);
    imshow("Distance Transform Image", dist);
  6. We threshold the dist image and then perform some morphology operation (i.e. dilation) in order to extract the peaks from the above image:
    // Threshold to obtain the peaks
    // This will be the markers for the foreground objects
    threshold(dist, dist, .4, 1., CV_THRESH_BINARY);
    // Dilate a bit the dist image
    Mat kernel1 = Mat::ones(3, 3, CV_8UC1);
    dilate(dist, dist, kernel1);
    imshow("Peaks", dist);
  7. From each blob then we create a seed/marker for the watershed algorithm with the help of the cv::findContours function:
    // Create the CV_8U version of the distance image
    // It is needed for findContours()
    Mat dist_8u;
    dist.convertTo(dist_8u, CV_8U);
    // Find total markers
    vector<vector<Point> > contours;
    // Create the marker image for the watershed algorithm
    Mat markers = Mat::zeros(dist.size(), CV_32SC1);
    // Draw the foreground markers
    for (size_t i = 0; i < contours.size(); i++)
    drawContours(markers, contours, static_cast<int>(i), Scalar::all(static_cast<int>(i)+1), -1);
    // Draw the background marker
    circle(markers, Point(5,5), 3, CV_RGB(255,255,255), -1);
    imshow("Markers", markers*10000);
  8. Finally, we can apply the watershed algorithm, and visualize the result:
    // Perform the watershed algorithm
    watershed(src, markers);
    Mat mark = Mat::zeros(markers.size(), CV_8UC1);
    markers.convertTo(mark, CV_8UC1);
    bitwise_not(mark, mark);
    // imshow("Markers_v2", mark); // uncomment this if you want to see how the mark
    // image looks like at that point
    // Generate random colors
    vector<Vec3b> colors;
    for (size_t i = 0; i < contours.size(); i++)
    {
    int b = theRNG().uniform(0, 255);
    int g = theRNG().uniform(0, 255);
    int r = theRNG().uniform(0, 255);
    colors.push_back(Vec3b((uchar)b, (uchar)g, (uchar)r));
    }
    // Create the result image
    Mat dst = Mat::zeros(markers.size(), CV_8UC3);
    // Fill labeled objects with random colors
    for (int i = 0; i < markers.rows; i++)
    {
    for (int j = 0; j < markers.cols; j++)
    {
    int index = markers.at<int>(i,j);
    if (index > 0 && index <= static_cast<int>(contours.size()))
    dst.at<Vec3b>(i,j) = colors[index-1];
    else
    dst.at<Vec3b>(i,j) = Vec3b(0,0,0);
    }
    }
    // Visualize the final image
    imshow("Final Result", dst);
cv::RNG::uniform
int uniform(int a, int b)
returns uniformly distributed integer random number from [a,b) range
cv::Vec3b
Vec< uchar, 3 > Vec3b
Definition: matx.hpp:364
cv::Mat::rows
int rows
the number of rows and columns or (-1, -1) when the matrix has more than 2 dimensions
Definition: mat.hpp:1959
cv::watershed
void watershed(InputArray image, InputOutputArray markers)
Performs a marker-based image segmentation using the watershed algorithm.
cv::filter2D
void filter2D(InputArray src, OutputArray dst, int ddepth, InputArray kernel, Point anchor=Point(-1,-1), double delta=0, int borderType=BORDER_DEFAULT)
Convolves an image with the kernel.
cv::theRNG
RNG & theRNG()
Returns the default random number generator.
cv::NORM_MINMAX
@ NORM_MINMAX
flag
Definition: base.hpp:196
cv::cvtColor
void cvtColor(InputArray src, OutputArray dst, int code, int dstCn=0)
Converts an image from one color space to another.
cv::distanceTransform
void distanceTransform(InputArray src, OutputArray dst, OutputArray labels, int distanceType, int maskSize, int labelType=DIST_LABEL_CCOMP)
Calculates the distance to the closest zero pixel for each pixel of the source image.
cv::Mat::at
_Tp & at(int i0=0)
Returns a reference to the specified array element.
cv::threshold
double threshold(InputArray src, OutputArray dst, double thresh, double maxval, int type)
Applies a fixed-level threshold to each array element.
cv::waitKey
int waitKey(int delay=0)
Waits for a pressed key.
CV_BGR2GRAY
@ CV_BGR2GRAY
Definition: types_c.h:121
CV_CHAIN_APPROX_SIMPLE
@ CV_CHAIN_APPROX_SIMPLE
Definition: types_c.h:458
CV_8U
#define CV_8U
Definition: interface.h:67
CV_32SC1
#define CV_32SC1
Definition: interface.h:106
CV_THRESH_BINARY
@ CV_THRESH_BINARY
Definition: types_c.h:571
cv::Mat::convertTo
void convertTo(OutputArray m, int rtype, double alpha=1, double beta=0) const
Converts an array to another data type with optional scaling.
CV_32F
#define CV_32F
Definition: interface.h:72
cv::dilate
void dilate(InputArray src, OutputArray dst, InputArray kernel, Point anchor=Point(-1,-1), int iterations=1, int borderType=BORDER_CONSTANT, const Scalar &borderValue=morphologyDefaultBorderValue())
Dilates an image by using a specific structuring element.
cv::imread
Mat imread(const String &filename, int flags=IMREAD_COLOR)
Loads an image from a file.
cv::Vec
Template class for short numerical vectors, a partial case of Matx.
Definition: matx.hpp:306
cv::Mat::cols
int cols
Definition: mat.hpp:1959
CV_8UC3
#define CV_8UC3
Definition: interface.h:84
CV_RETR_EXTERNAL
@ CV_RETR_EXTERNAL
Definition: types_c.h:446
cv::Mat::size
MatSize size
Definition: mat.hpp:1978
uchar
unsigned char uchar
Definition: interface.h:47
CV_RGB
#define CV_RGB(r, g, b)
Definition: imgproc_c.h:985
cv::imshow
void imshow(const String &winname, InputArray mat)
Displays an image in the specified window.
cv::drawContours
void drawContours(InputOutputArray image, InputArrayOfArrays contours, int contourIdx, const Scalar &color, int thickness=1, int lineType=LINE_8, InputArray hierarchy=noArray(), int maxLevel=INT_MAX, Point offset=Point())
Draws contours outlines or filled contours.
cv::Point
Point2i Point
Definition: types.hpp:183
cv::datasets::index
@ index
Definition: gr_skig.hpp:77
cv::Mat
n-dimensional dense array class
Definition: mat.hpp:741
cv::Matx::cols
@ cols
Definition: matx.hpp:92
i
for i
Definition: modelConvert.m:63
CV_DIST_L2
@ CV_DIST_L2
Definition: types_c.h:559
CV_8UC1
#define CV_8UC1
Definition: interface.h:82
cv
Definition: affine.hpp:52
cv::bitwise_not
void bitwise_not(InputArray src, OutputArray dst, InputArray mask=noArray())
Inverts every bit of an array.
cv::Mat_< float >
cv::findContours
void findContours(InputOutputArray image, OutputArrayOfArrays contours, OutputArray hierarchy, int mode, int method, Point offset=Point())
Finds contours in a binary image.
cv::normalize
static Vec< _Tp, cn > normalize(const Vec< _Tp, cn > &v)
cv::datasets::circle
@ circle
Definition: gr_skig.hpp:62
cv::Mat::data
uchar * data
pointer to the data
Definition: mat.hpp:1961
CV_THRESH_OTSU
@ CV_THRESH_OTSU
Definition: types_c.h:577