how to get adaptive contrast enhancement transformation of the gray image ?

can this be done without using adapthisteq()?

the local variation of the lesions obtained by

g=255((sigmoid(color)-sigmoid(graymin))/(sigmoid(graymax) - sigmoid(graymin)) where # * sigmoid(color) = 1/(1+exp((mean - colorimage)/variance) # * sigmoid(graymin) is the min sigmoid value of the grayimage # * sigmoid(graymax) is the max sigmoid value of the grayimage

how can i get colorimage subtracted from meanimage which is grayscale

Answer 1

[Image Analyst]

Sure, you can do almost whatever you want if you write it yourself. You can make a color image out of a grayscale image and then subtract:

grayImage3D = cat(3, grayImage, grayImage, grayImage);
% Cast to float to avoid clipping at 0.
newImage = double(rgbImage) - double(grayImage3D);  % It's floating point now.

[kalaivaani]

how can i obtain the maximum and minimum intensity value of greenchannel image and its value is in 1*700 size but it should be subtracted from meanimage of the greenchannel image with size 605*700 what should i do

[Image Analyst]

minValue = min(grayImage(:));
maxValue = max(grayImage(:));
subtractedImage = grayImage - uint8(minValue);

[kalaivaani]

fmax = 1./(1+exp((double(meanimage) - $)./(double(varianceimage)));
fmin = 1./(1+exp((double(meanimage) - $)./(double(varianceimage)));

were $ should be replaced by maximum and minimum value of greenchannel image(size of 1*700) respectively and mean and variance are greenchannel image with size (605*700)if i replace it as you given ??? Error using ==> minus Class of operand is not supported.

Error in ==> Untitled2 at 30
fmax = 1./(1+exp(((meanimage) - uint8(Gmax)))./(varianceimage));

what should i do , please correct it.

[Image Analyst]

I have no idea what Gmax, varianceimage, or meanimage are, so I don't know how to correct it, other than meanimage must be a uint8, or else don't use int8 around Gmax.

[kalaivaani]

Gmax is the maximum intensity value of greenchannel image(1*700) varianceimage &meanimage are greenchannel images (605*700)

??? Undefined function or method 'exp' for input arguments of type
'uint8'.

Error in ==> Untitled2 at 30
fmax = 1./(1+exp((uint8(meanimage)
-uint8(Gmax)))./uint8(varianceimage));

[Image Analyst]

You gotta learn to read the error messages. It says it doesn't want uint8, so what do you think you should do now?