OpenCV Image Segmentation

Image segmentation refers to the process of segmenting digital image into multiple image subregions, and realizes three kinds of algorithms related to image segmentation in OpenCV, which are: Watershed segmentation algorithm, pyramid segmentation algorithm and mean drift segmentation algorithm.

Watershed Segmentation algorithm
Watershed segmentation algorithms require you or a previous algorithm to provide a tag that specifies which areas are targets, which are roughly the background, and so on; the segmentation effect of watershed segmentation algorithm relies heavily on the tokens provided. The algorithm is implemented by the function cvwatershed in OpenCV

Pyramid Segmentation algorithm
The Pyramid segmentation algorithm is implemented by Cvprysegmentation, the function is simple to use, the size of the image and the number of layers of the pyramid, the width and height of the image must be divisible by 2, and the number of times divisible by 2 determines the maximum number of layers of the pyramid.

Mean Drift segmentation algorithm
The mean drift segmentation algorithm is realized by cvprymeanshiftfiltering, and the mean drift segmentation pyramid layer can only be between [1,7]

Friendship links, personal feeling better in this respect blog:

Http://www.cnblogs.com/xrwang/archive/2010/02/28/ImageSegmentation.html