The image calculation found that they are all black or all white.

The image calculation found that they are all black or all white.

my coding

img1= imread('SCM Data111.jpg');
img = (-0.18 / (-0.28 / (45.39 /img1 - 1))+1) * 5.3;
imwrite(img, 'n_.jpg');

Can you tell me the reason why the subsequent pictures are all black and how to solve it?

Answer 1

[Walter Roberson]

img = (-0.18 / (-0.28 / (45.39 /double(img1) - 1))+1) * 5.3;

When you do calculations with integer datatypes, the results of the calculations are converted to the integer data type.

[bozheng]

Unfortunately, the pictures that caused the error are all black.

[Walter Roberson]

Your formula divides by the input, but parts of the input can be 0 (especially where there is black.) That leads to large output values in places -- but also leads to small output values for a lot of the image because the division by the large-valued components comes out small.

filename = 'https://www.mathworks.com/matlabcentral/answers/uploaded_files/1521731/image.jpeg';
img1 = imread(filename);
min(img1(:)), max(img1(:))

ans = uint8 0

ans = uint8 255

img = (-0.18 / (-0.28 / (45.39 /double(img1) - 1))+1) * 5.3;

min(img(:)), max(img(:))

ans = 2.4993

ans = Inf

syms x
Q = @(v) sym(v);
imgs = (-Q(0.18) / (-Q(0.28) / (Q(45.39) /x - 1))+1) * Q(5.3)

imgs = 

fplot(imgs, [0 255])

ir = uint8(rescale(img, 0, 255, 'InputMin', 2.5, 'InputMax', 50));
imshow(ir); colorbar

[bozheng]

can you tell me why we know the value is 'InputMin', 2.5, 'InputMax', 50 thanks

Answer 2

[Image Analyst]

img1 is an array so you need to use dot division

img = (-0.18 ./ (-0.28 ./ (45.39 ./ img1 - 1)) + 1) * 5.3;

And double check your parentheses to make sure they're correct. For example is

(45.39 ./ img1 - 1)

supposed to be

((45.39 ./ img1) - 1)

or

(45.39 ./ (img1 - 1))