Databases
StackOverflow2013
Postsdbo
POOpenCV using k-means to posterize an image
 

Note that there are some explanatory texts on larger screens.

plurals
  1. POOpenCV using k-means to posterize an image
    primarykey
    data
    text
    <p>I want to posterize an image with k-means and OpenCV in C++ interface (cv namespace) and I get weird results. I need it for reduce some noise. This is my code:</p> <pre><code>#include "cv.h" #include "highgui.h" using namespace cv; int main() { Mat imageBGR, imageHSV, planeH, planeS, planeV; imageBGR = imread("fruits.jpg"); imshow("original", imageBGR); cv::Mat labels, data; cv::Mat centers(8, 1, CV_32FC1); imageBGR.convertTo(data, CV_32F); cv::kmeans(data, 8, labels, cv::TermCriteria(CV_TERMCRIT_ITER, 10, 1.0), 3, cv::KMEANS_PP_CENTERS, &amp;centers); imshow("posterized hue", data); data.convertTo(data, CV_32FC3); waitKey(); return 0; } </code></pre> <p>But I get a weird result</p> <p><img src="https://i.stack.imgur.com/dGy2i.png" alt="Fruit"></p> <p>First image: original</p> <p>Second image: after k-means.</p> <p>Any advice?</p> <hr> <h1>Update: the right solution. maybe someone can help me in optimize the code?</h1> <pre><code>#include "cv.h" #include "highgui.h" #include &lt;iostream&gt; using namespace cv; using namespace std; int main() { Mat src; src = imread("fruits.jpg"); imshow("original", src); blur(src, src, Size(15,15)); imshow("blurred", src); Mat p = Mat::zeros(src.cols*src.rows, 5, CV_32F); Mat bestLabels, centers, clustered; vector&lt;Mat&gt; bgr; cv::split(src, bgr); // i think there is a better way to split pixel bgr color for(int i=0; i&lt;src.cols*src.rows; i++) { p.at&lt;float&gt;(i,0) = (i/src.cols) / src.rows; p.at&lt;float&gt;(i,1) = (i%src.cols) / src.cols; p.at&lt;float&gt;(i,2) = bgr[0].data[i] / 255.0; p.at&lt;float&gt;(i,3) = bgr[1].data[i] / 255.0; p.at&lt;float&gt;(i,4) = bgr[2].data[i] / 255.0; } int K = 8; cv::kmeans(p, K, bestLabels, TermCriteria( CV_TERMCRIT_EPS+CV_TERMCRIT_ITER, 10, 1.0), 3, KMEANS_PP_CENTERS, centers); int colors[K]; for(int i=0; i&lt;K; i++) { colors[i] = 255/(i+1); } // i think there is a better way to do this mayebe some Mat::reshape? clustered = Mat(src.rows, src.cols, CV_32F); for(int i=0; i&lt;src.cols*src.rows; i++) { clustered.at&lt;float&gt;(i/src.cols, i%src.cols) = (float)(colors[bestLabels.at&lt;int&gt;(0,i)]); // cout &lt;&lt; bestLabels.at&lt;int&gt;(0,i) &lt;&lt; " " &lt;&lt; // colors[bestLabels.at&lt;int&gt;(0,i)] &lt;&lt; " " &lt;&lt; // clustered.at&lt;float&gt;(i/src.cols, i%src.cols) &lt;&lt; " " &lt;&lt; // endl; } clustered.convertTo(clustered, CV_8U); imshow("clustered", clustered); waitKey(); return 0; } </code></pre> <p>Result:</p> <p><img src="https://i.stack.imgur.com/QBVcV.png" alt="Posterized Fruit"></p>
    singulars
    1. PO
      singulars
      1. PTAnswer
    1. This table or related slice is empty.
    1. PTQuestion
    1. USBill the Lizard
    1. USnkint
    plurals
    1. This table or related slice is empty.
    1. PL
      singulars
      1. LTLinked
    1. This table or related slice is empty.
    1. PO
      singulars
      1. PTAnswer
    2. PO
      singulars
      1. PTAnswer
    1. VO
      singulars
      1. POOpenCV using k-means to posterize an image
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. POOpenCV using k-means to posterize an image
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. POOpenCV using k-means to posterize an image
      1. This table or related slice is empty.
      1. VTUpMod
    1. COIt could just be that you need more iterations and/or a smaller epsilon. I'd suggest you try removing "CV_TERMCRIT_EPS" for now, and play with the number of iterations in your TermCriteria. See if that helps.
      singulars
      1. POOpenCV using k-means to posterize an image
      1. USjilles de wit
    2. COi'm just approaching computer vision, image processing and machine learning but for me there is another mistake in what i'm doing and not only a parameter fine tuning..
      singulars
      1. POOpenCV using k-means to posterize an image
      1. USnkint
    3. COI'm not suggesting you start parameter tuning, I'm suggesting you simplify your code to test that what you're trying works in its most basic form. Removing epsilon and increasing the number of iterations removes unneeded trickery.
      singulars
      1. POOpenCV using k-means to posterize an image
      1. USjilles de wit
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload