How to detect contour self-intersection with C++ and OpenCV? - c++

Question

I need to test contour on self-intersection but I don't know how it implement. Or how I can detect only contours without self-intersection in cv::Mat?
F.ex. left contour must be matched, right contour don't matched

Answer 1

Here is a solution:
Skeleton + pruning => reduce the contours to a single pixel width
For each pixel, compute the number of neighbors
If a pixel has more than 2 neighbors, then there it is in the middle of an intersection.
(optional) Connected component labeling in order to separate the different shapes.
You can also use a Hough transform.

Answer 2

If the lines are represented by a polygon (you know the corner points), you may draw the lines on an accumulation matrix.
Declare an new blank cv::Mat of type CV_8UC1 and initialize it with zero values. For every pixel between the two lines, increment the matrix by 1.
I am not if using the cv::line method is the best way to accomplish this task (you may create a new image for every line and sum up all the images as the final step). The best way that I can think of is to increment the points by using the equation of the line.
When you draw lines that intersect, in the accumulation matrix you'll have values of 2. If you find them, you'll know that the contour has self-intersections and you also know where they are.
If you have the image as an input, then the previously mentioned solution might work.
Best regards!

Answer 3

I tried ma best to implement it but couldn't due to lack the logic to code it. The logic i tried is you have the set of points of contours. Now check the occurrence of each point i.e how many number of times each point has appeared, if it has appeared more then one time it indicates the intersection point.
Let me know if i'm wrong.
the code i tried isn't working for this logic maybe someone might help you with it.
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <iostream>
#include "opencv2/imgproc/imgproc.hpp"
using namespace cv;
using namespace std;
RNG rng(12345);
int main( )
{
Mat image;
image = imread("0.png", CV_LOAD_IMAGE_COLOR); // Read the file
if(! image.data ) // Check for invalid input
{
cout << "Could not open or find the image" << std::endl ;
return -1;
}
cvtColor( image, image, CV_BGR2GRAY );
namedWindow( "Display window12", WINDOW_AUTOSIZE );// Create a window for display.
imshow( "Display window12", image );
Mat drawing;
vector<vector<Point> > contours;
vector<Vec4i> hierarchy;
findContours( image, contours, hierarchy, CV_RETR_TREE, CV_CHAIN_APPROX_SIMPLE, Point(0, 0) );
int m = 1;
vector<Point> contours1;
for(int i= 0; i < contours.size(); i++)
{
for(int j= 0; j < contours[i].size();j++) // run until j < contours[i].size();
{
contours1.push_back(Point (contours[i][j]));
// cout << contours[i][j] << "contours1"<<contours1<<endl; //do whatever
}
}
cout<<contours.size();
// Finding the occrence of each point it has appeared
//for(int i=0;i<contours.size();i++)
//{
// for(int j=0; j<contours[i].size();j++) // run until j < contours[i].size();
// {
// //contours1.push_back(Point (contours[i][j]));
// //if (contours[i][j] == contours[i][j])
// if( contours[i] ==contours1.at(i).x)
// // if( posX ==points.at(p).x)
// cout<<"hi";
// // cout << contours[i][j] << "contours1"<<contours1<<endl; //do whatever
// }
//}
namedWindow( "Display window", WINDOW_AUTOSIZE );// Create a window for display.
imshow( "Display window", image );
waitKey(0); // Wait for a keystroke in the window
return 0;
}