Some questions about clustering using the kmeans function in matlab

Some questions about clustering using the kmeans function in matlab

Share with you

I am a beginner in image processing. When I was learning k-means clustering, I encountered some things I didn’t understand. I learned some experience from forums and blogs. I will make a record here to provide you with some ideas and conclusions. Encourage yourself. If there is any incompetence or copyright infringement, please leave a private message, and I will deal with it as soon as possible. Thank you all again, this is the first time I write a blog, and the writing and content are quite crude.

Kmeans function clustering method

At the beginning of learning k-means clustering, the function kmeans in matlab was learned from time to time. It was learned and debugged in the popular oopp moderator https://www.cnblogs.com/HelloDreams/p/5346849.html . Among them, the When rgb (image processing is mostly grayscale) image clustering, I want to change it to grayscale image clustering, and I encountered a problem here. Later, in the article
https://blog.csdn.net/justforacm/article/details/80941584#commentBox of the moderator of haocrab
and
https://blog.csdn.net/mingtian715/article/details/51534165 of Timmy_Y
, it was mentioned The following points are summarized:

1. kmeans is clustered in units of behavior
2. The unit of kmeans clustering should be the feature that can identify this point;
3. In the annotation of the kmeans function, X: N*P data matrix, N is the number of data, and P is a single data dimension ;