how to restore image after segmentation using watershed transform and otsu thresholding method?

I have segmented the image filled with noise using the watershed transform and otsu thresholding methods, then how do I restore the image to the original image?

[Rik]

I recovered the removed content from the Google cache (something which anyone can do). Editing away your question is very rude. Someone spent time reading your question, understanding your issue, figuring out the solution, and writing an answer. Now you repay that kindness by ensuring that the next person with a similar question can't benefit from this answer.

This page is now archived on the Wayback Machine.

Answer 1

[Image Analyst]

You need to save it. Don't destroy it by overwriting it with a labeled image from watershed or imbinarize. Then it will be there for you when you need it again. Otherwise if you overwrite it, there's no transforming a labeled or binary image back into a gray scale image. So that's why you need to save it and not destory it.

[Gold]

is there an image restoration algorithm that can return the segmentation results of the watershed and otsu threshold methods?

[Image Analyst]

You can "return the segmentation results of the watershed and otsu threshold methods? " just by accepting the outputs of those function calls. If you then want to return/pass them to other functions you can then do that. But that would not be called a "restoration algorithm", it's simply passing variables around.

[Gold]

how can the result be compared to the original image before segmentation? to remove watershed lines and the result of Otsu segmentation in the form of a binary image, while the original image is rgb/gray? please guide

[Image Analyst]

How do you want to "compare" them? You could display them side by side. You could use imfuse. You could use imoverlay to burn the lines into the original image. There is no one single definition of compare, and you didn't tell us your definition.

[Gold]

I tested an image with watershed and otsu methods. then I want to compare which method is the best. For that I need to change the image from the segmentation of the two methods to the initial image. then test it using SNR. so what's the solution? Thank you

[Image Analyst]

I'm sorry but that makes no sense to me. A watershed divides an image up into regions, which may have completely different intensities. Otsu thresholding finds regions brighter and darker than a certain global threshold. In general you could not use them both. They are not simply slightly different method of each other. They do completely different things. The solution depends on what you want to do or find in the image (which you're not showing us (the code or the image)).