how to set generalised threshold values in histogram eqivalent to get a clear image

I have a grayscale image. I've performed histeq() function on it. I have to write a generalised code by comparing pixels with its neighbouring pixels for different images to find threshold values to make the image much clearer. Ive attached grayscale image and its histeq image. I have to give threshold values for histeq function to get image much more clearer to detect tumour.

Answer 1

[Image Analyst]

Histogram equalization does not help with determining a threshold. Plus it gives crummy looking images. Don't waste your time.

[rohith bharadwaj]

Is there a better way to make the image clear in grayscale representation. So that I can easily distinguish tumour . It should be easy for me to do contour segmentation later.

[Image Analyst]

Some kind of spatial filtering may help, but just a global readjusting of graylevels like a histogram equalization does is useless. All that does is make the threshold at a different value than it would otherwise be at but it doesn't give you any different binary image than if you hadn't done equalization.