Best segmentation algorithm

Question

I am trying to develop a system that recognizes various objects present in an image based on their primitive functions, such as texture, shape and color.

The first step in this process is to extract individual objects from the image, and then process the images one by one.

However, the segmentation algorithm that I have studied so far is not even close to the perfect or so-called ideal image segmentation algorithm.

Segmentation accuracy will determine how much better the system responds to a given request.

Segmentation must be fast and accurate.

Can anyone suggest me some kind of segmentation algorithm, developed or implemented so far, which will not be too complicated to implement, but will be fair enough to complete my project.

Any help is available ..

Answer 1

A very late answer, but may help someone find this on google, as this question appeared as the first result for the “best segmentation algorithm”.

Fully convolutional networks seem to perform exactly the task you are asking for. Check out the arXiv document , and the implementation in MatConvNet .

CNN ( , , 3 , FCN-8).

Answer 2

, -. , . , , , , - . , (-) , --- * FFT_Face *.
. * Water_face *. /. * Water_Face * * FFT_Face image *. . , . , .

, , .

Answer 3

Google : http://www.cse.iitb.ac.in/~sharat/papers/prim.pdf

, , , , , , , .

Answer 4

you can try the watershed segmentation algorithm also you can calculate the accuracy of the segmentation algorithm with quality measures