Hello i want write a program to make a segmentation of fingerprint image

1) Divide the image into number of sub images of size 16x16 pixels. 2) The local mean for each sub image is calculated. 3) The local variance for each sub image is calculated. 4) The local coherence for all image 5) show Images notes\ image what i use (gray.bmp) please if can help me

[Walter Roberson]

What is your question? You appear to have the algorithm already. If you do not know the basics of programming in MATLAB then refer to the documentation Getting Started

Answer 1

[Abhishek Ballaney]

https://in.mathworks.com/discovery/image-segmentation.html

[alaa alanssari]

hello my friend thank you for your responding, my teacher asked me to work on (grey fingerprint image) to do the following: 1- divided image to 16*16 blocks. 2- calculate local mean for each blocks. 3- calculate local variance for each blocks. 4- calculate coherence for each blocks. 5- Aggregation all blocks in one image' 6- repeat (1,2,3,4) for image.

so i search in internet to find any way to help but i didn't finds i hope you can help me. thank you for your time and attention best wishes

Answer 2

[Image Analyst]

See my attached demos. They are for blockproc() which will process tiles of your image in a non-overlapping manner.

If you want the 16x16 tiles to overlap, so that the block only moves over by one pixel, then simply use conv2() and stdfilt() - a lot simpler than blockproc().