D-Alex
commented
11 years ago Hi,
the InputArray / OutputArrays of opencv are a bit tricky especially in combination with ruby. If a c++ function requests a InputArray or OutputArray as argument ropencv will always convert this argument to cv::Mat before passing it to the c++ function.
Unfortunately, if you use a std::Vector for an OutputArray this will fail because to convert a std::Vector to a cv::Mat a copy is created and ropencv does not copy the result back into the given std::vector (this will be fixed at some point).
Back to your problem: OpenCV is expecting an OutputArray to store the circles as Point3f. Therefore, it will be happy if you call it with a cv::Mat as argument.
circles = cv::Mat.new
cv::HoughCircles(gray, circles, CV_HOUGH_GRADIENT, 1, gray.rows/10, 10, 20 );
#reshape result --> each row stores one circle
circles = circles.reshape(1,circles.cols)
circles.each_row do |circle|
puts cv::Point.new(circle[0].round, circle[1].round)
end
bobbdelsol
based on the documentation, i think that Circles is supposed to be a vector
circles – Output vector of found circles. Each vector is encoded as a 3-element floating-point vector (x, y, radius) .
fitting it to cv.HoughCircles void Mat image Mat circles /O int method double dp double minDist double param1 100 double param2 100 int minRadius 0 int maxRadius 0
i get an error that the content doesn't type doesn't match or an error that the content is wrong.
circles.size if a ptr if you set it up as Mat,
of course i could just be confused as how to translate vector circles to ruby
my code
def test_hough hough_circles 'public/g/edit_images/stuff.jpg' end
def hough_circles(img_name)
Mat src, src_gray
/ Read the image
src = cv::imread(img_name, 1)
return -1 unless src.data src_gray=src
/ Convert it to gray
cv::cvtColor(src, src_gray, CV_BGR2GRAY)
/ Reduce the noise so we avoid false circle detection
arraysize=cv::Size.new arraysize.width=9 arraysize.height=9 cv::GaussianBlur(src_gray, src_gray, arraysize, 2, 2)
arraysize.width=3 arraysize.height=1 circles=cv::Mat.new(arraysize,cv::CV_64F)
circles=Vector::Mat.new
/ Apply the Hough Transform to find the circles
cv::HoughCircles(src_gray, circles, CV_HOUGH_GRADIENT, 1, src_gray.rows/8, 200, 100, 0, 0)
/ Draw the circles detected
(0..circles.size).each do |i| Point center(cvRound(circles[i][0]), cvRound(circles[i][1])) radius = cvRound(circles[i][2])
circle center
end