Introduction - If you have any usage issues, please Google them yourself
The K means algorithm for image segmentation, the input is a color image, convert to grayscale image segmentation, the output of grayscale images. The use of gray as the characteristics of each pixel clustering, due to light and other reasons, and sometimes should belong to an object pixel, its gray value will also be very different, may lead to clustering of the pixel error has occurred. in the segmentation results, the surface, there would be different from other pixel noise points, so , the algorithm Finally, the results of a median filter to eliminate noise, to the role of smoothing the image