What method can I use for background subtraction before performing analysis on my image?

Hi, i am exploring matlab methods for personal growth and i want to determine how to:

perform background substraction

contrast correction, gaussian filtering and thresholding and binarization for my image.

I have attached a demo m file and image and i get this error:

Unrecognized function or variable 'lowThreshold'.

Error in testfiltering (line 53)

mask = grayImage > lowThreshold & grayImage < highThreshold;

Answer 1

[Image Analyst]

You need to assign that value. See the interactive thresholding program in my File Exchange for how to do that.

See my Image Processing Tutorial in my File Exchange:

Image Analyst's File Exchange

[Neo]

@Image Analyst

I have been exploring your image processing tutorial. how do I know if the thresholidng i see matches the thresholding i want? I cannotconfirm by vision if the image has had background substraction, contrast correction, gaussian filtering and thresholding and binarization. I can just see that its changed. I adjusted the thresholding value in your tutorial to 50 (instead of 100 which was preset). What do you think?

[Image Analyst]

I don't know how to answer this. If you started with an image and a background image, you would certainly know if your code did a background subtraction. If you don't have the original image and the background image, then where did you get the image that you have?

"how do I know if the thresholidng i see matches the thresholding i want?" I don't know what this means. It MUST match the threshold you want. If it was not what you wanted, then why did you accept it? Obviously you're only going to accept a threshold that you want. You're not going to accept a threshold that you do not want, of course. That said, it may be that there is no threshold value that will give you a perfect segmentation just by thresholding alone.

[Neo]

@Image Analyst

i see. yes, this was not a well thought out response from me. i accept your answer and will adjust to my current code,