Localized Adaptive Contrast enhancement on opencv and python

Question

I am trying to do image enhancement on different types of images. I have tried a lot of different methods but usually, it works only on one image but fails on a different image. I am new to Image Processing so it difficult for me to understand How it needs to be done. I am posting the input images and desired output.

Input 1:

output 1:

Input 2:

output 2:

Input 3:

output 3:

Input 4:

Input4

Output 4:

I have tried CLAHE , Adaptive Thresholding , Histogram Equilization , Auto Brightness and Contrast adjustment , Adaptive Integral Enhancement , and many more things. Nothing is working on all the images. My assumption is I need to implement some kind of Local Adaptive Contrast Enhancement, But then again I may be wrong. It would be of a great help if anyone could help me out with this problem.

Answer 1

As fmw42 said, you potentially need to work on gray level images. So the suggestion was to convert to grayscale and invert as needed. I can suggest that you convert to Y Cb Br and apply on Y, and keep Cb Cr as is and see whether this works. This is the method I would use, say, to apply histogram equalization to a color image. CLAHE should do the trick for the gray level work.