why the imread output result is not same as the input of the imwrite?

if true % code data_test= [1 2 3 4; 2 5 0 6; 4 10 0 3; 5 8 6 14]; imwrite((data_test),'clown.png'); s=imread('clown.png') end run the code above,i get that s =

255  255  255  255
255  255    0  255
255  255    0  255
255  255  255  255

If i turn the data type of data_test to uint8,(imwrite(uint8(data_test),'clown.png');) the result of s is same as data_test. why the input of imwrite must turn to uint8? thank you very much.

Answer 1

[Guillaume]

When pixels are of type double, as is the cas for your data_test, intensities are assumed to be in the range 0-1. Anything below 0 is considered the same as 0, anything above 1 is considered to be 1.

Thus, when you save your data_test as double, it is just converted to: [1 1 1 1; 1 1 0 1;1 1 0 1; 1 1 1 1]. As PNG images don't support double pixels, it is then converted to uint8. Hence why, you get the result back as uint8.

Note that im2uint8 produces the same result:

im2uint8(data_test)

Now, the uint8 function is not designed for images but just for plain number, and will convert any number in the range 0-255 to unsigned 8 bit integer with the same value. Therefore there is no clipping. Compare the above with:

im2uint8(uint8(data_test))

In conclusion, if your pixels are indeed uint8, declare them as so, otherwise matlab will consider them to range from 0 to 1.